run_clonealign: Run clonealign across a range of initializations

In kieranrcampbell/clonealign: Assigning scRNA-seq to clone-of-origin using copy number from ultra-low-depth scDNA-seq

Description

Run clonealign across a range of initializations and select the fit that acheives the best evidence lower bound (ELBO).

Usage

run_clonealign(
  gene_expression_data,
  copy_number_data,
  initial_shrinks = c(0, 5, 10),
  n_repeats = 3,
  print_elbos = TRUE,
  ...
)

Arguments

gene_expression_data	A matrix of gene counts or a SingleCellExperiment. See ?clonealign
copy_number_data	A matrix or data frame of copy number calls for each clone. See ?clonealign.
initial_shrinks	Initial shrinkages for the clone assignment variational parameters
n_repeats	Number of fits to perform at each initial shrink
print_elbos	Logical - should the ELBOs inferred be printed?
...	Additional arguments to pass to clonealign(...)

Details

This function essentially wraps clonealign and can be interacted with as such. The parameter initial_shrinks controls how hard the variational parameters are initially assigned, analagous to the E-step in EM. At 0, they are initialized evenly across clones, while at 10 they are semi hard assigned to the most likely initial values.

Value

The clonealign_fit object for the fit that maximizes the ELBO. See ?clonealign for details.

Examples

library(SummarizedExperiment)
data("example_sce")
L <- rowData(example_sce)[,c("A", "B", "C")]
run_clonealign(example_sce, L, initial_shrinks = c(1), n_repeats = 1)

kieranrcampbell/clonealign documentation built on Dec. 18, 2020, 3:49 a.m.