The function runs trainCS_gene over the entire set of genes needed for deconvolution from the reference dataset.

trainCS(
  yref,
  yref_need,
  seed = 1218,
  method = c("lar", "lasso", "enet", "ridge", "l1", "TV", "l2"),
  par = T,
  n.cores,
  lambda = 0.1
)

Arguments

yref

matrix, numeric expression matrix for target genes

yref_need

matrix, numeric expression matrix of needed genes

seed

numeric, random seed

method

vector, character vector of optimization methods

par

logical, T/F for parallelization

n.cores

numeric, number of cores

lambda

numeric, penalty paramemter for non-linear optimization

Value

list with coefficients and predictive R2