network_sift {inferCSN}

network_sift

Description

network_sift

Usage

network_sift(
  network_table,
  matrix = NULL,
  meta_data = NULL,
  pseudotime_column = NULL,
  method = c("entropy", "max"),
  entropy_method = c("Shannon", "Renyi"),
  effective_entropy = FALSE,
  shuffles = 100,
  entropy_nboot = 300,
  history_length = 1,
  entropy_p_value = 0.05,
  cores = 1,
  verbose = TRUE
)

Arguments

network_table

network_table

matrix

The expression matrix.

meta_data

The meta data for cells or samples.

pseudotime_column

The column of pseudotime.

method

method The method used for filter edges. Could be choose "entropy" or "max".

entropy_method

If setting 'method' to 'entropy', could be choose "Shannon" or "Renyi" to compute entropy.

effective_entropy

Logical value, using effective entropy to filter weights or not.

shuffles

The number of shuffles used to calculate the effective transfer entropy. Default is 'shuffles' = 100.

entropy_nboot

entropy_nboot

history_length

history_length

entropy_p_value

P value.

cores

Number of CPU cores used. Setting to parallelize the computation with foreach.

verbose

Print detailed information.

Value

Filtered network table

Examples

data("example_matrix")
data("example_ground_truth")
network_table <- inferCSN(example_matrix)
network_table_filtered <- network_sift(network_table)
data("example_meta_data")
network_table_filtered_entropy <- network_sift(
  network_table,
  matrix = example_matrix,
  meta_data = example_meta_data,
  pseudotime_column = "pseudotime",
  history_length = 2,
  shuffles = 0,
  entropy_nboot = 0
)

network.heatmap(
  example_ground_truth[, 1:3],
  heatmap_title = "Ground truth",
  show_names = TRUE,
  rect_color = "gray70"
)
network.heatmap(
  network_table,
  heatmap_title = "Raw",
  show_names = TRUE,
  rect_color = "gray70"
)
network.heatmap(
  network_table_filtered,
  heatmap_title = "Filtered",
  show_names = TRUE,
  rect_color = "gray70"
)
network.heatmap(
  network_table_filtered_entropy,
  heatmap_title = "Filtered by entropy",
  show_names = TRUE,
  rect_color = "gray70"
)

auc.calculate(
  network_table,
  example_ground_truth,
  plot = TRUE
)
auc.calculate(
  network_table_filtered,
  example_ground_truth,
  plot = TRUE
)
auc.calculate(
  network_table_filtered_entropy,
  example_ground_truth,
  plot = TRUE
)