gene_importances: Calculate the importance of a feature

In SCORPIUS: Inferring Developmental Chronologies from Single-Cell RNA Sequencing Data

Calculate the importance of a feature

Description

Calculates the feature importance of each column in x in trying to predict the time ordering.

Usage

gene_importances(
  x,
  time,
  num_permutations = 0,
  ntree = 10000,
  ntree_perm = ntree/10,
  mtry = ncol(x) * 0.01,
  num_threads = 1,
  ...
)

Arguments

x	A numeric matrix or a data frame with M rows (one per sample) and P columns (one per feature).
time	A numeric vector containing the inferred time points of each sample along a trajectory as returned by infer_trajectory.
num_permutations	The number of permutations to test against for calculating the p-values (default: 0).
ntree	The number of trees to grow (default: 10000).
ntree_perm	The number of trees to grow for each of the permutations (default: ntree / 10).
mtry	The number of variables randomly samples at each split (default: 1% of features).
num_threads	Number of threads. Default is 1.
...	Extra parameters passed to ranger.

Value

a data frame containing the importance of each feature for the given time line

Examples

dataset <- generate_dataset(num_genes=500, num_samples=300, num_groups=4)
expression <- dataset$expression
group_name <- dataset$sample_info$group_name
space <- reduce_dimensionality(expression, ndim=2)
traj <- infer_trajectory(space)
# set ntree to at least 1000!
gene_importances(expression, traj$time, num_permutations = 0, ntree = 1000)

SCORPIUS documentation built on Aug. 8, 2023, 1:09 a.m.