simu_base_param: Given parameters of a reference dataset, calculate the ZINB...

In mqzhanglab/BSDE: Barycenter Single-Cell Differential Expression for Case-Control Studies

Given parameters of a reference dataset, calculate the ZINB parameters for simulating a new dataset of a different size.

Description

The basic parameters of ZINB(μ, θ, z) from the reference dataset can be estimated with DCA (Deep Count Autoencoder).

Usage

simu_base_param(
  t_mean,
  t_disp,
  t_drop,
  t_meta,
  nTotal = 30,
  nall = 40,
  RIN_adj = FALSE
)

Arguments

t_mean	mean parameter matrix μ of the reference dataset (gene x cell).
t_disp	dispersion parameter matrix θ of the reference dataset (gene x cell).
t_drop	dropout parameter matrix z of the reference dataset (gene x cell).
t_meta	Data frame (nrow = cell) of meta information. It must contain a column individual recording from which subject each cell is collected. If RIN_adj=TRUE, then it must also contain column RIN for RNA integrity number.
nTotal	numbers of genes to simulate
nall	numbers of individuals to simulate
RIN_adj	If TRUE, then t_meta must contain column RIN for RNA integrity number. (default: FALSE)

Value

a list containing:

• sample_ctrl: gene x individual x param array, which gives the parameter μ, θ and z of ZINB for simulation.

• RNA.simu: residual standard deviations of means.

References

Eraslan, G., Simon, L. M., Mircea, M., Mueller, N. S., & Theis, F. J. (2019). Single-cell RNA-seq denoising using a deep count Autoencoder. Nature Communications, 10(1), 1-14.

mqzhanglab/BSDE documentation built on March 20, 2022, 5:17 a.m.