Unified Interface for SVG Detection

A unified interface to run different spatially variable gene (SVG) detection methods. This function provides a consistent API for all supported methods.

Usage

CalSVG(
  expr_matrix,
  spatial_coords,
  method = c("meringue", "seurat", "binspect", "sparkx", "nnsvg", "markvario"),
  n_threads = 1L,
  verbose = TRUE,
  ...
)

Arguments

expr_matrix: Numeric matrix of gene expression values. Rows are genes, columns are spatial locations (spots/cells). Should be normalized (e.g., log-transformed counts).
spatial_coords: Numeric matrix of spatial coordinates. Rows are spatial locations, columns are x and y (and optionally z) coordinates. Row names should match column names of expr_matrix.
method: Character string specifying the SVG detection method. One of: "meringue", "seurat", "binspect", "sparkx", "nnsvg", "markvario".
n_threads: Integer. Number of threads for parallel computation. Default is 1. Set to higher values for faster computation on multi-core systems.
verbose: Logical. Whether to print progress messages. Default is TRUE.
...: Additional arguments passed to the specific method function.

Value

A data.frame containing SVG detection results. The exact columns depend on the method used, but typically include:

gene: Gene identifiers
pval or p.value: Raw p-values
padj or p.adj: Adjusted p-values (multiple testing corrected)
Method-specific statistics (e.g., Moran's I, LR statistic, odds ratio)

Details

This function serves as a wrapper around the individual method functions:

method = "meringue": Calls CalSVG_MERINGUE
method = "seurat": Calls CalSVG_Seurat
method = "binspect": Calls CalSVG_binSpect
method = "sparkx": Calls CalSVG_SPARKX
method = "nnsvg": Calls CalSVG_nnSVG
method = "markvario": Calls CalSVG_MarkVario

For method-specific parameters, please refer to the documentation of individual method functions.

Examples

# \donttest{
# Simulate example data
set.seed(42)
n_genes <- 20
n_spots <- 100
expr_matrix <- matrix(rpois(n_genes * n_spots, lambda = 10),
                      nrow = n_genes, ncol = n_spots)
rownames(expr_matrix) <- paste0("gene_", seq_len(n_genes))
colnames(expr_matrix) <- paste0("spot_", seq_len(n_spots))

spatial_coords <- cbind(x = runif(n_spots, 0, 100),
                        y = runif(n_spots, 0, 100))
rownames(spatial_coords) <- colnames(expr_matrix)

# Run SPARK-X method (no external dependencies)
results <- CalSVG(expr_matrix, spatial_coords, method = "sparkx",
                  kernel_option = "single", verbose = FALSE)
head(results)
#>      gene      p.value        p.adj stat_linear  pval_linear
#> 1  gene_9 4.440892e-16 3.195435e-14    5.688811 5.153065e-16
#> 2  gene_1 5.773160e-15 1.970518e-13    5.616795 5.713384e-15
#> 3 gene_13 8.215650e-15 1.970518e-13    5.925955 8.188767e-15
#> 4 gene_15 2.005129e-11 3.606967e-10    4.080433 2.005131e-11
#> 5  gene_7 7.587642e-11 1.091934e-09    3.971108 7.587649e-11
#> 6 gene_20 1.442864e-10 1.730349e-09    4.718315 1.442863e-10
# }