Frechet Single-Cell Distance (FSD)

Computes the single-cell analogue of Frechet Inception Distance (FID) on low-dimensional PCA embeddings between reference and simulated cell populations.

Usage

calc_frechet_singlecell_distance(
  ref_mat,
  sim_mat,
  cells_as_cols = TRUE,
  n_pcs = 15,
  max_features = 2000
)

Arguments

ref_mat

Matrix of reference cells (cells x features or features x cells).

sim_mat

Matrix of simulated cells (cells x features or features x cells).

cells_as_cols

Logical, whether cells are columns (default TRUE).

n_pcs

Number of principal components to evaluate (default 15).

max_features

Number of top variable features to retain (default 2000).

Value

A list with Frechet distance (FSD), mean discrepancy, and covariance trace discrepancy.

Examples

ref <- matrix(stats::rnorm(100), 50, 2)
sim <- matrix(stats::rnorm(100), 50, 2)
calc_frechet_singlecell_distance(ref, sim)
#> $fsd
#> [1] 11.88083
#> 
#> $fsd_squared
#> [1] 141.1541
#> 
#> $mean_discrepancy
#> [1] 60.01298
#> 
#> $cov_discrepancy
#> [1] 81.1411
#>