Signac

https://github.com/stuart-lab/signac
Admin
Signac is a comprehensive R package for analyzing single-cell chromatin data, offering quality...
Tokens:7,375
Snippets:47
Trust Score:6.4
Update:7 months ago
Show doc for...
Tokens:
Raw
### Install Signac from GitHub (R)

Source: https://github.com/stuart-lab/signac/blob/master/README.md

Installs the latest development version of the Signac R package directly from its GitHub repository. This method requires the 'remotes' package and is useful for users who want to access the newest features or contribute to development.

```r
if (!requireNamespace("remotes", quietly = TRUE))
    install.packages("remotes")
remotes::install_github("stuart-lab/signac", ref = "develop")
```

--------------------------------

### Install Signac from CRAN (R)

Source: https://github.com/stuart-lab/signac/blob/master/README.md

Installs the latest stable release of the Signac R package from CRAN. It requires setting repositories to include Bioconductor for automatic installation of dependencies. This is the recommended method for most users.

```r
setRepositories(ind=1:3) # needed to automatically install Bioconductor dependencies
install.packages("Signac")
```

--------------------------------

### Install Signac R Package

Source: https://github.com/stuart-lab/signac/blob/master/index.md

This code snippet demonstrates how to install the Signac R package from CRAN. It first sets the necessary repositories and then proceeds with the package installation. Ensure you have R and its package management system set up.

```r
setRepositories(ind=1:3)
install.packages("Signac")
```

--------------------------------

### Compute TSS Enrichment Score with TSSEnrichment (R)

Source: https://context7.com/stuart-lab/signac/llms.txt

Computes the Transcription Start Site (TSS) enrichment score per cell, a quality control metric for ATAC-seq data. It requires gene annotations to be set and can be run faster with `fast = TRUE`. The results are stored in the object's metadata.

```r
library(Signac)

# Compute TSS enrichment score
# Requires gene annotations to be set
pbmc <- TSSEnrichment(
  object = pbmc,
  fast = TRUE,        # Faster computation, less memory
  verbose = TRUE
)

# Results stored in metadata
head(pbmc$TSS.enrichment)
# [1] 6.234 8.123 5.876 7.234 9.012 4.567

head(pbmc$TSS.percentile)
# [1] 0.45 0.72 0.38 0.62 0.89 0.23

# Filter low-quality cells
pbmc <- subset(pbmc, subset = TSS.enrichment > 4)

# Visualize TSS enrichment distribution
VlnPlot(pbmc, features = "TSS.enrichment", pt.size = 0.1)
```

--------------------------------

### Call Peaks using MACS2

Source: https://context7.com/stuart-lab/signac/llms.txt

Calls peaks using MACS2 from fragment files, supporting group-specific peak calling. This function can identify cell type-specific regulatory regions. It requires the 'Signac' package and optionally MACS2 installed.

```r
library(Signac)

# Call peaks on all cells (pseudobulk)
peaks <- CallPeaks(
  object = pbmc,
  assay = "peaks",
  macs2.path = NULL,  # Auto-detect MACS2
  broad = FALSE,      # Narrow peaks for ATAC
  format = "BED",
  effective.genome.size = 2.7e9,  # Human genome
  extsize = 200,
  shift = -100,
  verbose = TRUE
)

# Call peaks per cell type
peaks <- CallPeaks(
  object = pbmc,
  assay = "peaks",
  group.by = "seurat_clusters",  # Call per cluster
  combine.peaks = TRUE,  # Merge overlapping peaks
  verbose = TRUE
)

# Returns GRanges object
head(peaks)
```

--------------------------------

### Run chromVAR for Motif Activity Scores

Source: https://context7.com/stuart-lab/signac/llms.txt

Runs the chromVAR algorithm to compute per-cell transcription factor motif activity scores. This requires motifs to be added to the assay beforehand and uses a genome reference. The results are stored in a new assay. Dependencies include 'Signac', 'chromVAR', and a BSgenome package.

```r
library(Signac)
library(chromVAR)
library(BSgenome.Hsapiens.UCSC.hg38)

# Run chromVAR (requires motifs to be added first)
pbmc <- RunChromVAR(
  object = pbmc,
  genome = BSgenome.Hsapiens.UCSC.hg38,
  assay = "peaks",
  new.assay.name = "chromvar"
)

# chromVAR scores stored as new assay
DefaultAssay(pbmc) <- "chromvar"

# Visualize TF activity
FeaturePlot(
  object = pbmc,
  features = c("MA0139.1", "MA0105.4"),  # CTCF, NFKB1
  min.cutoff = "q10",
  max.cutoff = "q90",
  pt.size = 0.1
)
```

--------------------------------

### Run SVD for LSI with RunSVD (R)

Source: https://context7.com/stuart-lab/signac/llms.txt

Runs Singular Value Decomposition (SVD) for Latent Semantic Indexing (LSI), a standard dimensional reduction method for scATAC-seq data. It takes a Signac object, specifies the assay and desired number of components, and can scale embeddings. The results are stored in a specified reduction slot.

```r
library(Signac)

# Complete LSI workflow
pbmc <- RunTFIDF(pbmc)
pbmc <- FindTopFeatures(pbmc, min.cutoff = "q0")
pbmc <- RunSVD(
  object = pbmc,
  assay = "peaks",
  reduction.key = "LSI_",
  reduction.name = "lsi",
  n = 50,           # Number of components
  scale.embeddings = TRUE,
  verbose = TRUE
)

# Check correlation with sequencing depth
# First component often captures technical variation
DepthCor(pbmc)

# Access LSI embeddings
lsi_embeddings <- Embeddings(pbmc, reduction = "lsi")
dim(lsi_embeddings)
# [1] 8849   50
```

--------------------------------

### Convert Genomic Coordinates with Signac Utilities

Source: https://context7.com/stuart-lab/signac/llms.txt

Provides utility functions for converting between string representations of genomic coordinates and GRanges objects. StringToGRanges parses coordinate strings into GRanges, while GRangesToString converts GRanges back to strings. These are essential for manipulating peak names and genomic regions.

```r
library(Signac)
library(GenomicRanges)

# Convert strings to GRanges
peak_strings <- c("chr1-10000-11000", "chr2-20000-21000", "chr3-30000-31000")

peaks_gr <- StringToGRanges(
  regions = peak_strings,
  sep = c("-", "-")  # Separator pattern
)

peaks_gr
# GRanges object with 3 ranges
#   seqnames    ranges strand
#      <Rle> <IRanges>  <Rle>
#   chr1    10000-11000      *
#   chr2    20000-21000      *
#   chr3    30000-31000      *

# Convert GRanges to strings
peak_names <- GRangesToString(
  grange = peaks_gr,
  sep = c("-", "-")
)

peak_names
```

--------------------------------

### Create ChromatinAssay Object in R

Source: https://context7.com/stuart-lab/signac/llms.txt

Creates a ChromatinAssay object to store single-cell chromatin data, including peak counts and fragment file references. This function is foundational for scATAC-seq analysis within Signac and requires Seurat and Signac libraries.

```r
library(Signac)
library(Seurat)

# Read 10x Genomics peak matrix
counts <- Read10X_h5(filename = "filtered_peak_bc_matrix.h5")

# Read cell metadata
metadata <- read.csv(
  file = "singlecell.csv",
  header = TRUE,
  row.names = 1
)

# Create ChromatinAssay with fragment file reference
chrom_assay <- CreateChromatinAssay(
  counts = counts,
  sep = c(":", "-"),  # Peak name separator (chr:start-end)
  fragments = "fragments.tsv.gz",
  min.cells = 10,     # Minimum cells per peak
  min.features = 200  # Minimum peaks per cell
)

# Create Seurat object
pbmc <- CreateSeuratObject(
  counts = chrom_assay,
  assay = "peaks",
  meta.data = metadata
)

# View ChromatinAssay information
pbmc[["peaks"]]
# ChromatinAssay data with 87561 features for 8849 cells
# Variable features: 0
# Genome:
# Annotation present: FALSE
# Motifs present: FALSE
# Fragment files: 1
```

--------------------------------

### Create Genome Browser Coverage Plots with Signac

Source: https://context7.com/stuart-lab/signac/llms.txt

Generates genome browser-style coverage plots showing accessibility signal across genomic regions. Can optionally display gene annotations, peak locations, and peak-to-gene links. Supports plotting for specific cell identities or clusters.

```r
library(Signac)
library(patchwork)

# Basic coverage plot
CoveragePlot(
  object = pbmc,
  region = "MS4A1",       # Gene name or coordinates
  features = "MS4A1",     # Show gene expression
  assay = "peaks",
  annotation = TRUE,      # Show gene annotations
  peaks = TRUE,           # Show peak locations
  links = TRUE,           # Show peak-gene links
  idents = c("CD4 Naive", "CD14 Mono", "B cells")
)

# Coverage plot for specific region
CoveragePlot(
  object = pbmc,
  region = "chr1-10000-20000",
  group.by = "seurat_clusters",
  extend.upstream = 1000,
  extend.downstream = 1000
)

# Combine multiple plots
p1 <- CoveragePlot(pbmc, region = "MS4A1")
p2 <- CoveragePlot(pbmc, region = "CD3D")
p1 | p2
```

--------------------------------

### Perform TF Footprinting Analysis with Signac

Source: https://context7.com/stuart-lab/signac/llms.txt

Performs transcription factor footprinting analysis by computing normalized Tn5 insertion frequency around motif instances. Requires Signac and a BSgenome, and allows specifying motifs, window sizes, and whether to consider only motifs within peaks.

```r
library(Signac)
library(BSgenome.Hsapiens.UCSC.hg38)

# Compute footprints for specific TFs
pbmc <- Footprint(
  object = pbmc,
  genome = BSgenome.Hsapiens.UCSC.hg38,
  motif.name = c("MA0139.1", "MA0105.4"),  # CTCF, NFKB1
  upstream = 250,
  downstream = 250,
  in.peaks = TRUE,  # Only motifs within peaks
  verbose = TRUE
)

# Extract footprint data for plotting
footprint_data <- GetFootprintData(
  object = pbmc,
  features = "MA0139.1",
  group.by = "seurat_clusters"
)

# Plot footprints
PlotFootprint(
  object = pbmc,
  features = "MA0139.1",
  group.by = "seurat_clusters"
)
```

--------------------------------

### Create Density Scatter Plot for QC Metrics (R)

Source: https://context7.com/stuart-lab/signac/llms.txt

Generates a scatter plot with density coloring to visualize the relationship between two QC metrics (e.g., nCount_peaks and TSS.enrichment) in a Signac object. The function can optionally display quantile lines and log-transform the x-axis. It is useful for identifying potential filtering thresholds.

```r
library(Signac)

# Visualize QC metrics with density coloring
DensityScatter(
  object = pbmc,
  x = "nCount_peaks",
  y = "TSS.enrichment",
  log_x = TRUE,
  quantiles = TRUE  # Show quantile lines
)

# Custom thresholds visualization
DensityScatter(
  object = pbmc,
  x = "nCount_peaks",
  y = "nucleosome_signal",
  log_x = TRUE
)
```

--------------------------------

### Create Fragment Object in R

Source: https://context7.com/stuart-lab/signac/llms.txt

Creates a Fragment object to store and validate fragment file information, including MD5 hashes and cell barcode mapping. This function is part of the Signac package and is used to manage fragment data for downstream analyses.

```r
library(Signac)

# Create fragment object from file path
fpath <- system.file("extdata", "fragments.tsv.gz", package = "Signac")

# Define cells and their mapping (optional name remapping)
cells <- colnames(x = atac_small)
names(x = cells) <- paste0("test_", cells)

# Create and validate fragment object
frags <- CreateFragmentObject(
  path = fpath,
  cells = cells,
  validate.fragments = TRUE,  # Check cells exist in file
  verbose = TRUE,
  tolerance = 0.5  # Allow 50% of cells to be missing
)

# Add to existing assay
Fragments(atac_small) <- frags

# Count fragments per cell
counts <- CountFragments(fragments = fpath)
head(counts)
#              CB frequency_count mononucleosome nucleosome_free reads_count
# 1 AAACGAAAGAAC           12345           4521            7824       24690
# 2 AAACGAAAGACC            8765           3214            5551       17530
```

--------------------------------

### Add Transcription Factor Motifs to Assay

Source: https://context7.com/stuart-lab/signac/llms.txt

Adds transcription factor motif information to a ChromatinAssay, creating a peak-by-motif matrix. This process requires motif position frequency matrices (PFMs) and a genome reference. Dependencies include 'Signac', 'JASPAR2020', 'TFBSTools', and a BSgenome package.

```r
library(Signac)
library(JASPAR2020)
library(TFBSTools)
library(BSgenome.Hsapiens.UCSC.hg38)

# Get position frequency matrices from JASPAR
pfm <- getMatrixSet(
  x = JASPAR2020,
  opts = list(
    collection = "CORE",
    tax_group = "vertebrates",
    all_versions = FALSE
  )
)

# Add motif information to assay
pbmc <- AddMotifs(
  object = pbmc,
  genome = BSgenome.Hsapiens.UCSC.hg38,
  pfm = pfm,
  assay = "peaks",
  verbose = TRUE
)

# Access motif information
motifs <- Motifs(pbmc)
GetMotifData(pbmc, slot = "motif.names")[1:5]
```

--------------------------------

### Create Feature Matrix from Genomic Regions

Source: https://context7.com/stuart-lab/signac/llms.txt

Quantifies fragments in specified genomic regions to create a feature-by-cell matrix. This is useful for generating count matrices from custom peak sets or re-quantifying existing peaks. Requires 'Signac' and 'GenomicRanges' packages.

```r
library(Signac)
library(GenomicRanges)

# Get fragment object
frags <- Fragments(pbmc)

# Define custom regions
custom_peaks <- GRanges(
  seqnames = c("chr1", "chr1", "chr2"),
  ranges = IRanges(
    start = c(10000, 20000, 15000),
    end = c(11000, 21000, 16000)
  )
)

# Quantify fragments in regions
counts <- FeatureMatrix(
  fragments = frags,
  features = custom_peaks,
  cells = colnames(pbmc),
  process_n = 2000,  # Regions per batch
  verbose = TRUE
)

# Create new assay from custom peaks
pbmc[["custom"]] <- CreateChromatinAssay(
  counts = counts,
  fragments = frags
)
```

--------------------------------

### Calculate Fragment Fraction in Regions with FractionCountsInRegion (R)

Source: https://context7.com/stuart-lab/signac/llms.txt

Calculates the fraction of ATAC-seq fragments that map to a specified set of genomic regions. This is commonly used for quality control, such as computing blacklist region ratios. It can utilize built-in blacklists or custom GRanges objects.

```r
library(Signac)

# Calculate fraction of reads in blacklist regions
pbmc$blacklist_ratio <- FractionCountsInRegion(
  object = pbmc,
  assay = "peaks",
  regions = blacklist_hg38_unified  # Built-in blacklist
)

# Custom region calculation
custom_regions <- GRanges(
  seqnames = c("chr1", "chr2"),
  ranges = IRanges(start = c(1, 1), end = c(1000000, 1000000))
)

pbmc$custom_fraction <- FractionCountsInRegion(
  object = pbmc,
  assay = "peaks",
  regions = custom_regions
)

# Filter cells with high blacklist ratio
pbmc <- subset(pbmc, subset = blacklist_ratio < 0.01)
```

--------------------------------

### Compute Depth Correlations with DepthCor (R)

Source: https://context7.com/stuart-lab/signac/llms.txt

Computes the correlation between LSI (Latent Semantic Indexing) components and sequencing depth. This is used to identify and potentially exclude technical variation components from downstream analyses. The output visualizes correlations per component.

```r
library(Signac)

# Compute depth correlations
depth_cor <- DepthCor(
  object = pbmc,
  reduction = "lsi",
  n = 30  # Number of components to check
)

# Visual output shows correlation per component
# First component typically has strong correlation (>0.9)
# and should be excluded from downstream analysis

# Use components 2:30 for downstream analysis
pbmc <- RunUMAP(
  object = pbmc,
  reduction = "lsi",
  dims = 2:30  # Exclude first component
)

pbmc <- FindNeighbors(
  object = pbmc,
  reduction = "lsi",
  dims = 2:30
)

pbmc <- FindClusters(
  object = pbmc,
  verbose = FALSE,
  algorithm = 3  # SLM algorithm
)
```

--------------------------------

### Link Peaks to Genes using Signac

Source: https://context7.com/stuart-lab/signac/llms.txt

Links peaks to genes based on correlation between peak accessibility and gene expression. Requires paired scATAC-seq and scRNA-seq data. It allows specifying the correlation method, distance window, and various cutoffs.

```r
library(Signac)

# Link peaks to genes (requires RNA assay)
pbmc <- LinkPeaks(
  object = pbmc,
  peak.assay = "peaks",
  expression.assay = "RNA",
  peak.slot = "counts",
  expression.slot = "data",
  method = "pearson",
  distance = 500000,      # 500kb window around gene
  min.distance = NULL,    # No minimum distance
  min.cells = 10,
  pvalue_cutoff = 0.05,
  score_cutoff = 0.05,
  verbose = TRUE
)

# Access links
links <- Links(pbmc)
head(links)
# GRanges object with peak-gene links

# Get peaks linked to specific genes
linked_peaks <- GetLinkedPeaks(
  object = pbmc,
  features = c("MS4A1", "CD3D"),
  min.abs.score = 0.1
)

# Get genes linked to specific peaks
linked_genes <- GetLinkedGenes(
  object = pbmc,
  features = "chr1-100000-101000",
  min.abs.score = 0.1
)
```

--------------------------------

### Identify Top Features with FindTopFeatures (R)

Source: https://context7.com/stuart-lab/signac/llms.txt

Identifies the most frequently observed features (peaks) based on total counts, used for feature selection before dimensional reduction. It allows filtering by a minimum cutoff, such as a quantile (e.g., 'q75' for top 25%) or a specific count.

```r
library(Signac)

# Select top 25% of peaks (q75 cutoff)
pbmc <- FindTopFeatures(
  object = pbmc,
  min.cutoff = "q75"  # Top 25% peaks
)

# Alternative: minimum cell cutoff
pbmc <- FindTopFeatures(
  object = pbmc,
  min.cutoff = 10  # Peaks in at least 10 cells
)

# Use all features
pbmc <- FindTopFeatures(
  object = pbmc,
  min.cutoff = "q0"  # All peaks
)

# Check variable features
length(VariableFeatures(pbmc))
# [1] 65670

# Feature metadata includes count and percentile rank
head(pbmc[["peaks"]][[]])
#                      count percentile
# chr1-10100-10442        23      0.012
# chr1-14970-15591       156      0.089
```

--------------------------------

### Compute Gene Activity Matrix and Normalize Data

Source: https://context7.com/stuart-lab/signac/llms.txt

Calculates gene activity scores from ATAC-seq data, adds them as a new assay, and then normalizes the data. This process involves defining features, upstream/downstream extensions, and normalization methods. It requires the 'Signac' package.

```r
gene.activities <- GeneActivity(
  object = pbmc,
  assay = "peaks",
  features = NULL,      # All protein-coding genes
  extend.upstream = 2000,   # Include 2kb promoter
  extend.downstream = 0,
  biotypes = "protein_coding",
  max.width = 500000,   # Exclude very long genes
  verbose = TRUE
)

# Add as new assay and normalize
pbmc[["RNA"]] <- CreateAssayObject(counts = gene.activities)

pbmc <- NormalizeData(
  object = pbmc,
  assay = "RNA",
  normalization.method = "LogNormalize",
  scale.factor = median(pbmc$nCount_RNA)
)

# Visualize marker genes
DefaultAssay(pbmc) <- "RNA"
FeaturePlot(
  object = pbmc,
  features = c("MS4A1", "CD3D", "LEF1", "NKG7", "TREM1", "LYZ"),
  pt.size = 0.1,
  max.cutoff = "q95",
  ncol = 3
)
```

--------------------------------

### Find Differentially Active TFs using Signac

Source: https://context7.com/stuart-lab/signac/llms.txt

Identifies differentially active transcription factors between cell populations using the FindMarkers function. It utilizes Wilcoxon rank-sum test and calculates average difference with a specified function.

```r
differential_activity <- FindMarkers(
  object = pbmc,
  ident.1 = "CD4 Naive",
  ident.2 = "CD8 Naive",
  test.use = "wilcox",
  mean.fxn = rowMeans,
  fc.name = "avg_diff"
)
```

--------------------------------

### Find Enriched Motifs in Differential Peaks

Source: https://context7.com/stuart-lab/signac/llms.txt

Identifies transcription factor motifs that are significantly enriched in a set of differentially accessible peaks compared to a background set. This helps in identifying TFs associated with specific cell states or conditions. Requires 'Signac'.

```r
library(Signac)

# Find differentially accessible peaks
da_peaks <- FindMarkers(
  object = pbmc,
  ident.1 = "CD4 Naive",
  ident.2 = "CD14 Mono",
  only.pos = TRUE,
  test.use = "LR",
  min.pct = 0.05,
  latent.vars = "nCount_peaks"
)

# Get top peaks
top_da_peaks <- rownames(da_peaks[da_peaks$p_val_adj < 0.05, ])

# Find enriched motifs in these peaks
enriched_motifs <- FindMotifs(
  object = pbmc,
  features = top_da_peaks,
  background = 40000,  # Number of background peaks
  assay = "peaks",
  verbose = TRUE
)

# View results
head(enriched_motifs)
```

--------------------------------

### Add Gene Annotations to ChromatinAssay in R

Source: https://context7.com/stuart-lab/signac/llms.txt

Adds gene annotations to a ChromatinAssay object, which is crucial for gene activity scoring, TSS enrichment, and peak-to-gene linkage. Annotations can be loaded from AnnotationHub or GTF files using Signac and GenomicRanges libraries.

```r
library(Signac)
library(AnnotationHub)
library(GenomicRanges)

# Method 1: Load from AnnotationHub
ah <- AnnotationHub()
query(ah, "EnsDb.Hsapiens.v98")

ensdb_v98 <- ah[["AH75011"]]

# Extract gene annotations from EnsDb
annotations <- GetGRangesFromEnsDb(ensdb = ensdb_v98)

# Change to UCSC style chromosome names (chr1, chr2, etc.)
seqlevels(annotations) <- paste0("chr", seqlevels(annotations))
genome(annotations) <- "hg38"

# Add annotations to Seurat object
Annotation(pbmc) <- annotations

# Method 2: Load from GTF file directly
gtf <- rtracklayer::import("genes.gtf")
Annotation(pbmc) <- gtf

# Access existing annotations
gene_annotations <- Annotation(pbmc)
head(gene_annotations)
# GRanges object with 6 ranges and 5 metadata columns
```

--------------------------------

### Perform TF-IDF Normalization with RunTFIDF (R)

Source: https://context7.com/stuart-lab/signac/llms.txt

Performs Term Frequency-Inverse Document Frequency (TF-IDF) normalization, a standard method for scATAC-seq data. It corrects for both sequencing depth and peak accessibility frequency. The function supports multiple methods, with method 1 being recommended.

```r
library(Signac)

# Run TF-IDF normalization (Method 1 recommended)
pbmc <- RunTFIDF(
  object = pbmc,
  method = 1,       # log(TF) * log(IDF)
  scale.factor = 10000,
  verbose = TRUE
)

# Alternative methods available:
# method = 2: TF * log(IDF)
# method = 3: log(TF) * IDF
# method = 4: TF * IDF

# TF-IDF normalized data stored in 'data' slot
head(GetAssayData(pbmc, layer = "data")[1:3, 1:3])
```

--------------------------------

### Compute Nucleosome Signal in R

Source: https://context7.com/stuart-lab/signac/llms.txt

Computes the nucleosome signal score per cell based on fragment length distribution using the NucleosomeSignal function from the Signac package. This score helps assess chromatin quality and cell integrity, and can be used for filtering cells.

```r
library(Signac)

# Compute nucleosome signal score
pbmc <- NucleosomeSignal(object = pbmc)

# Score stored in metadata
head(pbmc$nucleosome_signal)
# [1] 0.8234 1.2341 0.9876 1.0234 0.7654 1.3456

# Visualize fragment length distribution grouped by signal strength
pbmc$nucleosome_group <- ifelse(
  pbmc$nucleosome_signal > 4,
  "NS > 4",
  "NS < 4"
)

FragmentHistogram(object = pbmc, group.by = "nucleosome_group")

# Filter cells with poor nucleosome banding pattern
pbmc <- subset(pbmc, subset = nucleosome_signal < 4)
```

--------------------------------

### Compute Gene Activity Scores with GeneActivity (R)

Source: https://context7.com/stuart-lab/signac/llms.txt

Computes gene activity scores by counting ATAC-seq fragments within gene body and promoter regions. This function creates a proxy for gene expression from scATAC-seq data, which can be used for cell type annotation. It requires a Signac object and gene annotations.

```r
library(Signac)

# Compute gene activity scores
pbmc <- GeneActivity(
  object = pbmc,
  assay = "peaks",
  ranges = gene_annotation_granges,
  process_regions = TRUE
)

# Gene activity scores stored in the 'genebody' assay
head(GetAssayData(pbmc, assay = "genebody")[1:3, 1:3])
```

--------------------------------

### Find Closest Genomic Features using Signac

Source: https://context7.com/stuart-lab/signac/llms.txt

Finds the closest gene or genomic feature to a set of query regions. This function is useful for annotating peaks with their nearest genes or identifying features near specific genomic coordinates. It takes a list of regions as input.

```r
library(Signac)

# Find closest genes to peaks
closest <- ClosestFeature(
  object = pbmc,
  regions = rownames(pbmc)[1:100]  # First 100 peaks
)

head(closest)
#   gene_id   gene_name gene_biotype closest_region    query_region distance
# 1 ENSG0001  SAMD11    protein_coding chr1-10000-11000 chr1-9500-10500   500
# 2 ENSG0002  NOC2L     protein_coding chr1-15000-16000 chr1-14000-15000  1000

# Find closest peaks to specific coordinates
query_regions <- c("chr1-100000-100500", "chr2-50000-50500")
closest <- ClosestFeature(
  object = pbmc,
  regions = query_regions
)
```

--------------------------------

### Identify Accessible Peaks with Signac

Source: https://context7.com/stuart-lab/signac/llms.txt

Identifies peaks that are accessible (non-zero counts) in a specified set of cells. This function is useful for identifying cell type-specific regulatory regions. It allows filtering by cell identity or specific cell IDs and setting a minimum cell count.

```r
library(Signac)

# Find accessible peaks in specific clusters
accessible <- AccessiblePeaks(
  object = pbmc,
  assay = "peaks",
  idents = c("CD4 Naive", "CD8 Naive"),
  min.cells = 10  # Peak must be accessible in at least 10 cells
)

length(accessible)
# [1] 45678

# Find accessible peaks in specific cells
accessible <- AccessiblePeaks(
  object = pbmc,
  assay = "peaks",
  cells = WhichCells(pbmc, idents = "B cells"),
  min.cells = 5
)
```
Tokens:
Raw