get_DE_MRF: Run the MRF model to estimate posterior probabilities of...

In eddiehli/MRFscRNAseq: What the Package Does (One Line, Title Case)

Description

Run the MRF model to estimate posterior probabilities of differential expression for each gene across each cell type

Usage

get_DE_MRF(
  data,
  g_g,
  c_c,
  nulltype = 1,
  df = 15,
  iterEM = 200,
  iterGibbsPost = 20000,
  brPost = 10000
)

Arguments

g_g	Gene to gene network matrix
c_c	Cell to cell dependency matrix
nulltype	Type of null hypothesis assumed in estimating f0, see locfdr package.Default is the MLE (nulltype=1)
df	Degrees of freedom for fitting the estimated density, see locfdr package. Default df=15
iterEM	Max number of iterations for the EM algorithm. Default=200
iterGibbsPost	Number of Gibbs posterior samples. Default=20,000
brPost	Number of burn-in for the posterior samples. Default=10,000
zz	Summary statistics matrix, rows are genes, columns are cell types

Value

The estimated model parameters and the posterior probabilities of differential expression

postDE	Posterior probabilities of differential expression. A 2-dimensional array: (num of genes)*(num of cell types)
paraMRF	Estimated model parameters
paraMRFTrace	Trace of the estimated model parameters in the EM algorithm
paraVar	Variance-covariance matrix of the estimated model parameters in the EM algorithm

eddiehli/MRFscRNAseq documentation built on Dec. 30, 2021, 9:08 a.m.