Skip to contents

Calculate VISION scores (with auto-correlation scores)

vision_w_autocor_sc.Rd

Calculates VISION-type scores for pathways based on DeTomaso, et al. Compared to other score types, you can also calculate delta-type scores between positive and negative gene indices, think epithelial vs mesenchymal gene signature, etc. Additionally, this function also calculates the auto- correlation values, answering the question if a given signature shows non- random enrichment on the kNN graph. The kNN graph (and distance measures) will be generated on-the-fly based on the embedding you wish to use.

Usage

vision_w_autocor_sc(
  object,
  gs_list,
  embd_to_use,
  no_embd_to_use = NULL,
  use_knn = TRUE,
  vision_params = params_sc_vision(),
  streaming = NULL,
  random_seed = 42L,
  .verbose = TRUE
)

Arguments

object

SingleCells, MetaCells (or potentially other) class.

gs_list

Named nested list. Every element must itself be a list with at least a "pos" element holding the gene identifiers, and optionally a "neg" one. A bare character vector is not accepted. The gene identifiers need to be part of the variables of the object.

embd_to_use

String. The embedding to use. Whichever you chose, it needs to be part of the object.

no_embd_to_use

Optional integer. Number of embedding dimensions to use. If NULL all will be used.

use_knn

Boolean. Shall the internal kNN be used. If set to yes, you need to ensure consistency.

vision_params

List with vision parameters, see params_sc_vision() with the following elements:

  • n_perm - Integer. Number of random permutations

  • n_cluster - Integer. Number of random clusters to generate to associate each set with.

  • knn - List of kNN parameters. See params_knn_defaults() for available parameters and their defaults.

streaming

Optional Boolean. Shall the data be streamed in. Useful for larger data sets where you wish to avoid loading in the whole data. If NULL, will automatically detect. Ignored when applied to MetaCells.

random_seed

Integer. The random seed.

.verbose

Boolean or integer. Controls verbosity and returns run times. FALSE -> quiet, TRUE or 1L -> normal verbosity, 2L -> detailed verbosity.

Value

A list with the following elements:

  • vision_matrix - Matrix of cells x signatures with the VISION pathway scores as values.

  • auto_cor_dt - data.table with the auto-correlation results per gene set, i.e., auto_cor (1 - Gaery's C), p_val and fdr.

References

DeTomaso, et al., Nat. Commun., 2019

Examples

# scores plus whether they sit non-randomly on the kNN graph
sc <- demo_single_cells()
gs_list <- list(
  programme_a = list(pos = get_gene_names(sc)[1:10]),
  programme_b = list(pos = get_gene_names(sc)[21:30])
)
res <- vision_w_autocor_sc(
  sc,
  gs_list = gs_list,
  embd_to_use = "pca",
  vision_params = params_sc_vision(n_perm = 50L, n_cluster = 3L),
  .verbose = FALSE
)
res$auto_cor_dt
#>    gene_set_name  auto_cor      p_val        fdr
#>           <char>     <num>      <num>      <num>
#> 1:   programme_a 0.7248461 0.01960784 0.01960784
#> 2:   programme_b 0.7437420 0.01960784 0.01960784

unlink(sc@dir_data, recursive = TRUE, force = TRUE)