computeBanksy: Compute the component neighborhood matrices for the BANKSY...

In prabhakarlab/Banksy: Spatial transcriptomic clustering

Compute the component neighborhood matrices for the BANKSY matrix.

Description

Compute the component neighborhood matrices for the BANKSY matrix.

Usage

computeBanksy(
  se,
  assay_name,
  coord_names = NULL,
  compute_agf = FALSE,
  k_geom = 15,
  spatial_mode = c("kNN_median", "kNN_r", "kNN_rn", "kNN_rank", "kNN_unif", "rNN_gauss"),
  n = 2,
  sigma = 1.5,
  alpha = 0.05,
  k_spatial = 100L,
  M = NULL,
  sample_size = NULL,
  sample_renorm = TRUE,
  seed = NULL,
  dimensions = "all",
  center = TRUE,
  verbose = TRUE
)

Arguments

se	A SpatialExperiment, SingleCellExperiment or SummarizedExperiment object. If not a SpatialExperiment object, argument coord_names must be provided.
assay_name	A string scalar specifying the name of the assay to use.
coord_names	A string vector specifying the names in colData corresponding to spatial coordinates.
compute_agf	A logical scalar specifying whether to compute the AGF.
k_geom	An integer scalar specifying the number of neighbors to use. Values \in [15,30] work well.
spatial_mode	A string scalar specifying the kernel for neighborhood computation (default: kNN_median). kNN_median: k-nearest neighbors with median-scaled Gaussian kernel kNN_r: k-nearest neighbors with $1/r$ kernel kNN_rn: k-nearest neighbors with $1/r^n$ kernel kNN_rank: k-nearest neighbors with rank Gaussian kernel kNN_unif: k-nearest neighbors wth uniform kernel rNN_gauss: radial nearest neighbors with Gaussian kernel
n	A numeric scalar specifying the exponent of radius (for kNN_rn).
sigma	A numeric scalar specifying the std. dev. of Gaussian kernel (for rNN_gauss).
alpha	A numeric scalar specifying the radius used: larger alphas give smaller radii (for rNN_gauss).
k_spatial	An integer scalar specifying the initial number of neighbors to use (for rNN_gauss)
M	Advanced usage. A integer scalar specifying the highest azimuthal Fourier harmonic to compute. If specified, overwrites the use_agf argument.
sample_size	An integer scalar number of neighbors to sample from the neighborhood.
sample_renorm	A logical scalar specifying whether to renormalize the neighbor weights to 1.
seed	An integer scalar specifying seed for sampling the neighborhood.
dimensions	A character vector specifying the dimensions to use when computing neighborhood. subset of colnames of cell.locs allUses all colnames of spatialCoords to compute (default)
center	A logical scalar specifying whether to center higher order harmonics in local neighborhoods.
verbose	A logical scalar specifying verbosity.

Details

Given an expression matrix (as specified by assay_name), this function computes the mean neighborhood matrix (H0) and optionally, the azimuthal Gabor filter (AGF) matrix (H1). The number of neighbors used to define the spatial neighborhood is given by k_geom. Different kernels may be used to compute the neighborhood features, specified by spatial_mode.

Value

A SpatialExperiment / SingleCellExperiment / SummarizedExperiment object with neighborhood matrices added.

Examples

data(rings)
spe <- computeBanksy(rings, assay_name = "counts", M = 1, k_geom = c(15, 30))

prabhakarlab/Banksy documentation built on April 11, 2024, 2:45 p.m.