Archr-Local Skill

Comprehensive assistance with ArchR (Analysis of Regulatory Regions) - the premier toolkit for single-cell ATAC-seq data analysis, generated from official documentation.

When to Use This Skill

This skill should be triggered when:

Core Analysis Tasks

• Working with scATAC-seq data: Processing fragment files, creating Arrow files, quality control
• Dimensionality reduction and clustering: LSI analysis, UMAP visualization, cell clustering
• Peak analysis: Differential accessibility, co-accessibility, peak-to-gene links
• Motif enrichment: Finding regulatory motifs, transcription factor analysis
• Integration: Multiome analysis (scATAC + scRNA), batch correction
• Visualization: Browser tracks, embedding plots, TSS enrichment plots
• Project management: Creating projects, subsetting, doublet detection

Specific Use Cases

• Setting up new ArchR projects from fragment files or BAM files
• Creating UMAP embeddings and clustering scATAC-seq data
• Identifying cell clusters and markers
• Performing differential accessibility analysis
• Generating browser tracks for genomic regions
• Adding motif annotations and performing motif enrichment
• Integrating scATAC-seq with scRNA-seq data
• Exporting results and creating publication-ready plots

Questions That Need This Skill

• "How do I create an ArchR project from my scATAC-seq data?"
• "How do I identify cell types in my ATAC-seq data?"
• "How do I find differentially accessible peaks between clusters?"
• "How do I add motif information to my peaks?"
• "How do I visualize my data with UMAP plots?"

Quick Reference

Common Patterns

Pattern 1: Setting up ArchR and creating a project

library(ArchR)
addArchRGenome("hg38")
addArchRThreads(8)
addArchRLocking(locking = TRUE)
set.seed(1)

# Create Arrow files from fragment data
ArrowFiles <- createArrowFiles(
  inputFiles = atacFiles,
  sampleNames = names(atacFiles),
  minTSS = 4,
  minFrags = 1000,
  addTileMat = TRUE,
  addGeneScoreMat = TRUE
)

# Create ArchR project
proj <- ArchRProj(
  ArrowFiles = ArrowFiles,
  outputDirectory = "ArchR-Output",
  copyArrows = TRUE
)

Pattern 2: Quality control and filtering

# Plot TSS enrichment for QC
p <- plotTSSEnrichment(proj, groupBy = "Sample")
plotPDF(p, name = "TSS-Enrich", ArchRProj = proj)

# Filter doublets
proj <- filterDoublets(proj)

# Subset to high-quality cells
proj <- subsetArchRProject(
  ArchRProj = proj,
  cells = proj$cellColData[proj$cellColData$Clusters != "Doublet", ]
)

Pattern 3: Dimensionality reduction and clustering

# Add Iterative LSI
proj <- addIterativeLSI(
  ArchRProj = proj,
  useMatrix = "TileMatrix",
  name = "IterativeLSI",
  force = TRUE
)

# Add UMAP embedding
proj <- addUMAP(
  ArchRProj = proj,
  reducedDims = "IterativeLSI",
  name = "UMAP",
  nNeighbors = 30,
  minDist = 0.5,
  metric = "cosine"
)

# Add clusters
proj <- addClusters(
  input = proj,
  reducedDims = "IterativeLSI",
  method = "Seurat",
  name = "Clusters",
  resolution = 0.8
)

Pattern 4: Visualization and plotting

# Plot UMAP colored by sample
p1 <- plotEmbedding(ArchRProj = proj, colorBy = "cellColData", name = "Sample", embedding = "UMAP")

# Plot UMAP colored by clusters
p2 <- plotEmbedding(ArchRProj = proj, colorBy = "cellColData", name = "Clusters", embedding = "UMAP")

# Side-by-side comparison
ggAlignPlots(p1, p2, type = "h")
plotPDF(p1, p2, name = "Plot-UMAP-Sample-Clusters.pdf", ArchRProj = proj, addDOC = FALSE)

Pattern 5: Gene accessibility and marker analysis

# Add gene expression scores
proj <- addGeneScoreMatrix(
  ArchRProj = proj,
  useMatrix = "TileMatrix",
  matrixName = "GeneScoreMatrix"
)

# Plot gene scores on UMAP
p <- plotEmbedding(
  ArchRProj = proj,
  embedding = "UMAP",
  colorBy = "GeneScoreMatrix",
  name = "CD34",
  size = 1
)

Pattern 6: Differential accessibility analysis

# Identify marker peaks
markersPeaks <- getMarkerFeatures(
  ArchRProj = proj,
  useMatrix = "PeakMatrix",
  groupBy = "Clusters",
  bias = c("TSSEnrichment", "log10(nFrags)"),
  testMethod = "wilcoxon"
)

# Extract region matrix
markerList <- getFeatures(markersPeaks, name = "PeakMatrix")
heatmapPeaks <- plotMarkerHeatmap(
  peakMatrix = getMatrixFromProject(proj, useMatrix = "PeakMatrix"),
  features = markerList,
  cutOff = "FDR <= 0.1 & Log2FC >= 1"
)

Pattern 7: Co-accessibility analysis

# Add co-accessibility
proj <- addCoAccessibility(
  ArchRProj = proj,
  reducedDims = "IterativeLSI"
)

# Get co-accessibility loops
cA <- getCoAccessibility(
  ArchRProj = proj,
  corCutOff = 0.5,
  resolution = 1000,
  returnLoops = TRUE
)

# Plot browser tracks with co-accessibility loops
p <- plotBrowserTrack(
  ArchRProj = proj,
  groupBy = "Clusters",
  geneSymbol = c("CD14", "CD3D"),
  upstream = 50000,
  downstream = 50000,
  loops = cA
)

Pattern 8: Motif enrichment analysis

# Add motif annotations
proj <- addMotifAnnotations(ArchRProj = proj, motifSet = "cisbp", name = "Motif")

# Perform motif enrichment
enrichMotifs <- peakAnnoEnrichment(
  seMarker = getFeatures(markersPeaks, name = "PeakMatrix"),
  ArchRProj = proj,
  peakAnnotation = "Motif"
)

# Plot motif enrichment
plotEnrichHeatmap(enrichMotifs, cutOff = "FDR <= 0.1 & Log2FC >= 1")

Pattern 9: Multiome data integration

# Import scRNA-seq data
rnaMatrix <- import10xFeatureMatrix(inputFiles = rnaFiles)

# Add gene expression matrix to ArchR project
proj <- addGeneExpressionMatrix(
  input = proj,
  matrices = rnaMatrix,
  strictMatch = TRUE
)

# Create dimensionality reduction using both modalities
proj <- addCombinedDims(
  ArchRProj = proj,
  reducedDims = c("IterativeLSI", "GeneIntegrationMatrix"),
  name = "CombinedDims"
)

Pattern 10: Exporting results and data

# Export group BigWig files
bw <- getGroupBW(
  ArchRProj = proj,
  groupBy = "Clusters",
  normMethod = "ReadsInTSS",
  tileSize = 100
)

# Export group fragment files
frags <- getGroupFragments(
  ArchRProj = proj,
  groupBy = "Clusters"
)

# Save project
proj <- saveArchRProject(proj, outputDirectory = "ArchR-Project-Output", load = FALSE)

Key Concepts

Core ArchR Objects

• ArchRProject: Main container for single-cell ATAC-seq data and analyses
• ArrowFiles: Efficient storage format for fragment data
• PeakMatrix: Peak-by-cell accessibility matrix
• TileMatrix: Fixed-size tile-by-cell accessibility matrix
• GeneScoreMatrix: Gene-by-cell accessibility score matrix

Analysis Workflow

1. Data Input: Import fragment files/BAM files → Create Arrow files
2. Quality Control: TSS enrichment → Doublet filtering → Cell subsetting
3. Dimensionality Reduction: Iterative LSI → UMAP
4. Clustering: Identify cell populations
5. Downstream Analysis: Differential accessibility, motif analysis, co-accessibility

Important Parameters

• minTSS: Minimum TSS enrichment score for cell retention (usually 4-10)
• minFrags: Minimum fragments per cell (usually 1000-5000)
• resolution: Clustering resolution (higher = more clusters)
• corCutOff: Correlation cutoff for co-accessibility (usually 0.3-0.5)

Reference Files

This skill includes comprehensive documentation in references/:

Core Documentation

• getting_started.md - Installation, basic setup, and introduction to ArchR
• data_preparation.md - Input file formats, project creation, and data import
• dimensionality_reduction.md - LSI, UMAP, and other dimensionality reduction methods
• clustering.md - Cell clustering, group creation, and population identification
• visualization.md - Plotting functions and data visualization

Analysis Functions

• analysis_functions.md - Core analysis functions for modalities and integrative analysis
• peak_analysis.md - Peak calling, differential accessibility, and peak-to-gene linking
• gene_analysis.md - Gene scoring, expression integration, and gene-based analyses
• enrichment_analysis.md - GO analysis, motif enrichment, and pathway analysis
• trajectory_analysis.md - Pseudotime analysis, lineage trajectories, and differentiation

Advanced Topics

• advanced.md - Advanced techniques and specialized analyses
• integration.md - Multi-omics integration and batch correction
• export.md - Data export, result saving, and sharing

Utilities

• project_management.md - Project organization, subsetting, and management
• utility_functions.md - Helper functions and utilities
• visualization_functions.md - Additional plotting and visualization tools
• other.md - Miscellaneous topics and supplementary information

Reference Content Structure

Each reference file contains:

• Detailed explanations of functions and workflows
• Code examples with syntax highlighting
• Parameter descriptions and usage notes
• Best practices and troubleshooting tips

Use file names to navigate specific topics (e.g., view clustering.md for clustering guidance).

Working with This Skill

For Beginners

1. Start with getting_started.md - Learn ArchR installation and basic concepts
2. Read data_preparation.md - Understand how to format and import your data
3. Follow dimensionality_reduction.md and clustering.md - Create your first analyses
4. Use visualization.md - Learn to plot and interpret your results

For Intermediate Users

1. Review peak_analysis.md and gene_analysis.md - Perform differential analyses
2. Explore enrichment_analysis.md - Add regulatory insights to your work
3. Check integration.md - Combine multiple modalities or datasets
4. Use export.md - Generate publication-ready outputs

For Advanced Users

1. Study advanced.md - Implement sophisticated analytical techniques
2. Review trajectory_analysis.md - Study cellular differentiation
3. Customize utility_functions.md - Extend ArchR functionality
4. Optimize workflows using project_management.md

Navigation Tips

• Use specific file names in your queries (e.g., "show me clustering.md")
• Ask for specific functions by name (e.g., "how does addIterativeLSI work?")
• Request examples from particular documentation sections
• Use the Quick Reference patterns for common workflows

Resources

Documentation Structure

• references/ - Complete extracted documentation organized by topic
• Quick Reference - Frequently used code patterns and workflows
• Key Concepts - Essential terminology and best practices

Getting Help

• Reference specific files for detailed function descriptions
• Use Quick Reference patterns for common tasks
• Ask about specific parameters or troubleshooting scenarios

Notes

• ArchR is specifically designed for single-cell ATAC-seq data analysis
• The Arrow file format enables memory-efficient processing of large datasets
• ArchR integrates seamlessly with other Bioconductor packages
• All major scATAC-seq file formats are supported (10x, sci-ATAC, etc.)
• The toolkit includes extensive QC metrics and validation steps

Updating

To refresh this skill with updated documentation:

1. Re-run the scraper with the same configuration
2. This skill will be rebuilt with the latest information from the ArchR documentation