binSpect#

Last Updated: Jan 29, 2024

Description#

Previously: binGetSpatialGenes. BinSpect (Binary Spatial Extraction of genes) is a fast computational method that identifies genes with a spatially coherent expression pattern.

Usage#

binSpect(
  gobject,
  spat_unit = NULL,
  feat_type = NULL,
  bin_method = c("kmeans", "rank"),
  expression_values = c("normalized", "scaled", "custom"),
  subset_feats = NULL,
  subset_genes = NULL,
  spatial_network_name = "Delaunay_network",
  spatial_network_k = NULL,
  reduce_network = FALSE,
  kmeans_algo = c("kmeans", "kmeans_arma", "kmeans_arma_subset"),
  nstart = 3,
  iter_max = 10,
  extreme_nr = 50,
  sample_nr = 50,
  percentage_rank = 30,
  do_fisher_test = TRUE,
  adjust_method = "fdr",
  calc_hub = FALSE,
  hub_min_int = 3,
  get_av_expr = TRUE,
  get_high_expr = TRUE,
  implementation = c("data.table", "simple", "matrix"),
  group_size = "automatic",
  do_parallel = TRUE,
  cores = NA,
  verbose = T,
  knn_params = NULL,
  set.seed = NULL,
  bin_matrix = NULL,
  summarize = c("p.value", "adj.p.value")
)

Arguments#

gobject

giotto object

spat_unit

spatial unit

feat_type

feature type

bin_method

method to binarize gene expression

expression_values

expression values to use

subset_feats

only select a subset of features to test

subset_genes

deprecated, use subset_feats

spatial_network_name

name of spatial network to use (default = ‘spatial_network’)

spatial_network_k

different k’s for a spatial kNN to evaluate

reduce_network

default uses the full network

kmeans_algo

kmeans algorithm to use (kmeans, kmeans_arma, kmeans_arma_subset)

nstart

kmeans: nstart parameter

iter_max

kmeans: iter.max parameter

extreme_nr

number of top and bottom cells (see details)

sample_nr

total number of cells to sample (see details)

percentage_rank

percentage of top cells for binarization

do_fisher_test

perform fisher test

adjust_method

p-value adjusted method to use (see p.adjust)

calc_hub

calculate the number of hub cells

hub_min_int

minimum number of cell-cell interactions for a hub cell

get_av_expr

calculate the average expression per gene of the high expressing cells

get_high_expr

calculate the number of high expressing cells per gene

implementation

enrichment implementation (data.table, simple, matrix)

group_size

number of genes to process together with data.table implementation (default = automatic)

do_parallel

run calculations in parallel with mclapply

cores

number of cores to use if do_parallel = TRUE

verbose

be verbose

knn_params

list of parameters to create spatial kNN network

set.seed

set a seed before kmeans binarization

bin_matrix

a binarized matrix, when provided it will skip the binarization process

summarize

summarize the p-values or adjusted p-values

Details#

We provide two ways to identify spatial genes based on gene expression binarization. Both methods are identicial except for how binarization is performed.

  • 1. binarize: Each gene is binarized (0 or 1) in each cell with kmeans (k = 2) or based on rank percentile

  • 2. network: Alll cells are connected through a spatial network based on the physical coordinates

  • 3. contingency table: A contingency table is calculated based on all edges of neighboring cells and the binarized expression (0-0, 0-1, 1-0 or 1-1)

  • 4. For each gene an odds-ratio (OR) and fisher.test (optional) is calculated

Three different kmeans algorithmes have been implemented:

    1. kmeans: default, see kmeans

    1. kmeans_arma: from ClusterR, see KMeans_arma

  • 3. kmeans_arma_subst: from ClusterR, see KMeans_arma, but random subsetting the vector for each gene to increase speed. Change extreme_nr and sample_nr for control.

Other statistics are provided (optional):

  • Number of cells with high expression (binary = 1)

  • Average expression of each gene within high expressing cells

  • Number of hub cells, these are high expressing cells that have a user defined number of high expressing neighbors

By selecting a subset of likely spatial genes (e.g. soft thresholding highly variable genes) can accelerate the speed. The simple implementation is usually faster, but lacks the possibility to run in parallel and to calculate hub cells. The data.table implementation might be more appropriate for large datasets by setting the group_size (number of genes) parameter to divide the workload.

Value#

data.table with results (see details)