simulate_multinomial_loglikelihoods: Simulate draws from a multinomial distribution for various...

In zh542370159/dropSplit: Automatically identify cell-containing and empty droplets for droplet-based scRNAseq data using dropSplit

Simulate draws from a multinomial distribution for various values of N.

Description

Uses the approximation from Lun et al. ( https://www.biorxiv.org/content/biorxiv/early/2018/04/04/234872.full.pdf )

Usage

simulate_multinomial_loglikelihoods(
  profile_p,
  umis_per_bc,
  num_sims = 1000,
  jump = 1000,
  n_sample_feature_block = 1e+06,
  verbose = FALSE
)

Arguments

profile_p	Probability of observing each feature.
umis_per_bc	UMI counts per barcode (multinomial N).
num_sims	Number of simulations per distinct N value.
jump	Vectorize the sampling if the gap between two distinct Ns exceeds this.
n_sample_feature_block	Vectorize this many feature samplings at a time.
verbose	verbose. Default is FALSE.

Value

A list including:

• distinct_ns an array containing the distinct N values that were simulated.

• log_likelihoods a length(distinct_ns) x num_sims matrix containing the simulated log likelihoods.

zh542370159/dropSplit documentation built on June 19, 2022, 2:49 p.m.