Compute p value for DAR test between two cell types

Compute p value for DAR test between two cell types

Usage

DAR_by_LRT(
  pic_mat,
  capturing_rates,
  cell_type_labels,
  n_cores = 1,
  plen = NULL,
  min_frag_length = 25,
  max_frag_length = 600,
  estimation_approach = "MLE"
)

Arguments

pic_mat

The observed peak by cell PIC count matrix

capturing_rates

A vector of estimated capturing rates for each cell

cell_type_labels

A vector specifying cell type labels

n_cores

A numerical value to specify the number of cores in parallel, default = 1

plen

A vector of peak length

min_frag_length

The value for the s1 hyperparameter in the ssPoisson distribution, this stands for the minimum fragment length requirement such that the fragment can be amplifiable and mappable to genome. Default = 25

max_frag_length

The value for the s2 hyperparameter in the ssPoisson distribution, this stands for the max fragment length requirement such that the fragment can be amplifiable. Default = 600

estimation_approach

The approach for parameter estimation, either 'MLE' for condition 1 or 'ME' for condition 1+2. The 'MLE' approach is more accurate and usually it has a higher power, but it ignores the size filtering step in snATAC-seq data generation. Default is 'MLE'

Value

Log loss value