scmet_simulate: Simulate methylation data from scMET.

View source: R/scmet_simulate.R

scmet_simulateR Documentation

Simulate methylation data from scMET.

Description

General function for simulating datasets with diverse proprties. This for instance include, adding covariates X that explain differences in mean methylation levels. Or also defining the trend for the mean - overdispersion relationship.

Usage

scmet_simulate(
  N_feat = 100,
  N_cells = 50,
  N_cpgs = 15,
  L = 4,
  X = NULL,
  w_mu = c(-0.5, -1.5),
  s_mu = 1,
  w_gamma = NULL,
  s_gamma = 0.3,
  rbf_c = 1,
  cells_range = c(0.4, 0.8),
  cpgs_range = c(0.4, 0.8)
)

Arguments

N_feat

Total number of features (genomics regions).

N_cells

Maximum number of cells.

N_cpgs

Maximum number of CpGs per cell and feature.

L

Total number of radial basis functions (RBFs) to fit the mean-overdispersion trend. For L = 1, this reduces to a model that does not correct for the mean-overdispersion relationship.

X

Covariates which might explain variability in mean (methylation). If X = NULL, a 2-dim matrix will be generated, first column containing intercept term (all values = 1), and second colunn random generated covariates.

w_mu

Regression coefficients for covariates X. Should match number of columns of X.

s_mu

Standard deviation for mean parameter mu.

w_gamma

Regression coefficients of the basis functions. Should match the value of L. If NULL, random coefficients will be generated.

s_gamma

Standard deviation of dispersion parameter gamma.

rbf_c

Scale parameter for empirically computing the variance of the RBFs.

cells_range

Range (betwen 0 and 1) to randomly (sub)sample the number of cells per feature.

cpgs_range

Range (betwen 0 and 1) to randomly (sub)sample the number of CpGs per cell and feature.

Value

A simulated dataset and additional information for reproducibility purposes.

Examples

sim <- scmet_simulate(N_feat = 150, N_cells = 50, N_cpgs = 15, L = 4)


andreaskapou/scMET documentation built on June 1, 2022, 11:47 p.m.