Calculate spearman correlations between features.

Calculate spearman correlations between features in Seruat object. Sparse implementation enables faster calculation of spearman correlations without need to cast sparse expression matrix to dense matrix.

findCorMarkers(
  object,
  features.x = NULL,
  features.y = rownames(object),
  ncell.subset = 5000,
  geosketch.subset = F,
  assay = DefaultAssay(object),
  slot = "data",
  verbose = T
)

Arguments

object

Seurat object

features.x

feature or meta feature. Spearman correlation between features.x and features.y are computed.

features.y

feature or meta feature. Spearman correlation between features.x and features.y are computed.

ncell.subset

max number of cells to run analysis on. Default is 5000.

geosketch.subset

Use GeoSketch method to subsample scRNA-seq data while preserving rare cell states (https://doi.org/10.1016/j.cels.2019.05.003). Logical, T or F (Default F). Recommended if cell type representation is imbalanced.

assay

Assay to run spearman correlation on. Default is DefaultAssay(object).

slot

slot to run spearman correlation on. Default is data.

verbose

print progress. Default is T.

Value

data.frame with spearman correlations.

Author

Nicholas Mikolajewicz and Saket Choudhary (https://github.com/saketkc/blog/blob/main/2022-03-09/SparseSpearmanCorrelation.ipynb)