make.snm.obj: Make Snm Object

In Sage-Bionetworks/snm: Supervised Normalization of Microarrays

Description

Makes snm object and sets the values to those provided by the user. Does some quality control, but this should be expanded in the future.

Usage

make.snm.obj(Y, bio.var, adj.var, int.var, spline.dim, nbins, weights,diagnose,rm.adj)

Arguments

Y	Matrix of raw, log base 2 transformed data
bio.var	Model matrix of biological variables
adj.var	Model matrix of adjustment variables
int.var	Data frame with the different intensity-dependent variables presented along the columns
spline.dim	Dimension of basis spline used for modeling intensity-dependent effects
nbins	Number of bins to partition data into when estimating intensity-dependent effects
weights	Weights for each probe. Values unchanged by the algorithm. Used to down-weight influence of data when calculating intensity-dependent effects.
diagnose	A flag that determines if diagnostic plots should be produced.
rm.adj	If set to TRUE, all adjustment variables are removed from the normalized data. Set to FALSE if you intend to perform classical hypothesis on the normalized data.

Author(s)

Brig Mecham <[e-mail]> and John D. Storey <[e-mail]>

Examples

##---- Should be DIRECTLY executable !! ----
##-- ==>  Define data, use random,
##--	or do  help(data=index)  for the standard data sets.

## The function is currently defined as

function(Y, bio.var, adj.var,int.var,spline.dim, nbins,weights,diagnose,fast.snm,retBio) {
  if(colnames(bio.var)[1] == "(Intercept)") { bio.var <- bio.var[,-1] }
  else{ stop("Model matrix for biological variables must contain an Intercept term in the first column with the title (Intercept).  Please try again!")}
  snm.obj = list()
  snm.obj$n.arrays = ncol(Y)
  snm.obj$n.probes = nrow(Y)
  snm.obj$fast.snm = fast.snm
  snm.obj$retBio = retBio
  if(fast.snm) {
    snm.obj$adj.var <- cbind(rep(1,snm.obj$n.arrays))
    snm.obj$bio.var <- cbind(adj.var, bio.var)
    snm.obj$bio.var <- snm.obj$bio.var[,-1]
    snm.obj$user.adj.var <- adj.var
    snm.obj$user.bio.var <- bio.var
    snm.obj$df.full = ifelse(is.null(ncol(snm.obj$bio.var)), 1, ncol(snm.obj$bio.var)) + ncol(snm.obj$adj.var)
    snm.obj$df.null = ncol(snm.obj$adj.var)    
  }else{
    snm.obj$bio.var = bio.var
    snm.obj$adj.var = adj.var
    snm.obj$df.full = ifelse(is.null(ncol(bio.var)), 1, ncol(bio.var)) + ncol(adj.var)
    snm.obj$df.null = ncol(adj.var)
  }
  snm.obj$int.var = int.var
  snm.obj$individuals = NULL
  snm.obj$spline.dim = spline.dim
  snm.obj$nbins = nbins
  snm.obj$diagnose = diagnose
  snm.obj$dat=Y
  snm.obj$weights=weights
  snm.obj$r.dat=Y
  class(snm.obj) = "edge"
  snm.obj <- checkArguments(snm.obj)
  return(snm.obj)
}

Sage-Bionetworks/snm documentation built on May 9, 2019, 12:14 p.m.