Generates a simple synthetic, pseudo gene expression matrix

This is a function to create a synthetic, pseudo gene expression matrix with a pre-defined number of total genes, groups and number of genes differentially expressed in the groups. The function will create in a standard setting a 1000 x 90 matrix, and add optionally a small group into the mixture that is slightly more difficult to detect

Usage

synthetic_signal_matrix(
  no_grps = 3,
  per_group = 30,
  total_genes = 1000L,
  no_genes_up = 100L,
  add_small_group = TRUE,
  size_small_grp = 5L,
  seed = 10101L
)

Arguments

no_grps

Integer. Number of initial larger groups. Default: 3L.

per_group

Integer. Number of samples per larger groups. Default: 30L.

total_genes

Integer. Number of total genes in the matrix. Default: 1000L.

no_genes_up

Integer. Number of genes per group that are differentially expressed. Default: 100L.

add_small_group

Boolean. Add a smaller group that overlaps with group 1? Default: TRUE.

size_small_grp

Integer. Size of the smaller group.

seed

Integer. Initial random seed for generation of the synthetic data. Default: 10101L.

Value

A synthetic_matrix_simple class containing:

• mat - The random matrix

• diff - List of differentially expressed genes per group

• group - Look-up vector for sample to group mapping

Examples

# default 1000 x 90 matrix with three groups and a small fourth one
synthetic_gex <- synthetic_signal_matrix()
dim(synthetic_gex$mat)
#> [1] 1000   90
table(synthetic_gex$group)
#> 
#>      group1      group2      group3 small_group 
#>          25          30          30           5 
lengths(synthetic_gex$diff)
#>      group1      group2      group3 small_group 
#>         100         100         100         101