Wrapper function to generate bulk sparsification parameters
params_bulk_sparsity.RdParameters for simulate_dropouts(). Dropout falls
out of the library size rather than an explicit per-gene dropout curve: a
size factor s_j ~ LogNormal(0, capture_efficiency_sigma) is drawn per
sample, giving a target library size of target_library_size * s_j, and each
gene is binomially thinned towards that target.
Usage
params_bulk_sparsity(
strategy = "seq_depth",
target_library_size = 20000,
capture_efficiency_sigma = 0.5,
seed = 123L
)Arguments
- strategy
String. Which dropout strategy to apply. Currently only
"seq_depth". One of"seq_depth". Defaults to"seq_depth".- target_library_size
Numeric. Reference library size per sample. Defaults to
20000.0.- capture_efficiency_sigma
Numeric. Standard deviation of the LogNormal size-factor distribution. Larger values spread the library sizes further apart. Defaults to
0.5.- seed
Integer. Seed for reproducibility purposes. Defaults to
123L.
Value
A named list with the following elements:
strategy - String. Which dropout strategy to apply. Currently only
"seq_depth". One of"seq_depth". Defaults to"seq_depth".target_library_size - Numeric. Reference library size per sample. Defaults to
20000.0.capture_efficiency_sigma - Numeric. Standard deviation of the LogNormal size-factor distribution. Larger values spread the library sizes further apart. Defaults to
0.5.seed - Integer. Seed for reproducibility purposes. Defaults to
123L.