run_MRF: Run MRF

In biqing-zhu/MARBLES: What the Package Does (One Line, Title Case)

Run MRF

Description

Run the MRF model to estimate the DE state for each gene across cell types.

Usage

run_MRF(paraMRF, expr, c_c, x_init)

Arguments

paraMRF	Starting values of the model parameter \Phi.
expr	Pseudo-bulk expression list for a gene. Each item is a vector of the pseudo-bulk expression across all individuals for a cell type in a condition. The list names should start with 'Cond1_' or 'Cond2_', and followed by the name of the cell type.
c_c	Binary cell type relationship network matrix. 1 means connected, and 0 means not connected.
x_init	Initial DE status. Should be a binary vector of length nk. 1 means DE, and 0 means EE.

Value

The estimated model parameters and the DE status:

x_mat

Trace of the DE status for each cell type in the ICM algorithm. A 2-dimensional array: (num of cell types)*(num of iterations). The first column represents the initialization, and the last column represents the final DE status.

theta_mat

Trace of the estimated model parameters \Theta in the ICM algorithm. A 2-dimensional array: 2*(num of iterations). The first column represents the initialization, and the last column represents the final parameters.

phi_mat

Trace of the estimated model parameters \Phi in the ICM algorithm. A 2-dimensional array: 2*(num of iterations). The first column represents the initialization, and the last column represents the final parameters.

biqing-zhu/MARBLES documentation built on Dec. 9, 2024, 5:33 p.m.