R/global_norm.R

In ropensci/DoOR.functions: Integrating Heterogeneous Odorant Response Data into a Common Response Model: A DoOR to the Complete Olfactome

# normalize the merged data globally using weight
# 
# normalize the merged data globally using weight
# 
# Some data contribute more than others based on how many receptors and odors 
# were measured, so that we introduce a weighted mean to scale the model 
# response values. If all weights are equal, then the weighted mean is the same
# as the arithmetic mean. expression for weighted scale: ( 
# sum(weight_receptors[i]*(RMAX[i]/SMAX[i])) + 
# sum(weight_odors[i]*(RMAX[i]/SMAX[i])) ) / ( sum(weight_receptors[i]) + 
# sum(weight_odors[i]) ). The global normalized output is equal to that merged 
# consensus vector (in values [0, 1]) multiply by a weighted scale.
# 
# @param RMAX numeric vector, a vector across all studies of maximum responses 
#   for this receptor.
# @param SMAX numeric vector, a vector across all studies of maximum response in
#   all receptors available in each study.
# @param MV numeric vector, a vector of model response values to be normalized 
#   (for this receptor).
# @param name.Stud character vector, names of involved studies.
# @param weightGlobNorm data matrix, a receptor-by-study matrix, value 1 
#   indicates available measurement in this study for this receptor, value NA 
#   indicates NOT.
# @param responseRange data frame, the response range and how many odors were 
#   measured in each study.
# @author Shouwen Ma <\email{shouwen.ma@@uni-konstanz.de}>
# @examples
# 
# 
# rm(list=ls())
# library(DoOR.data)
# data(door_global_normalization_weights)
# data(door_response_range)
# mergedValues_ORx <- seq(0,1, length= 20) 		
# # merged response values of receptor (ORx).
# name.Stud <- c("Hallem.2006.EN", "Pelz.2006.ALEC50")
# RMAX <- c(100, 4)		# maximum responses of receptor (ORx) from study 
#         "Hallem.2006.EN" and "Pelz.2006.ALEC50" respectively.
# SMAX <- c(294,6.9)	# maximum recordings of all receptors from study 
#          Hallem.2006.EN" and "Pelz.2006.ALEC50" respectively.
# GN_values_ORx <- global_norm(RMAX, SMAX, MV = mergedValues_ORx, name.Stud, 
# weightGlobNorm = door_global_normalization_weights, 
# responseRange = door_response_range)
# 
global_norm <-
  function(RMAX,
     SMAX,
     MV,
     name.Stud,
     weightGlobNorm = door_default_values("door_global_normalization_weights"),
     responseRange = door_default_values("door_response_range")) {
    # take the data subset for weight that involves only given studies
    weight.frame <-
      data.frame(weightGlobNorm[, name.Stud])  			  
    weight_receptors <-
      apply(weight.frame, 2, function(x)
        length(which(x == 1))) # how many tested receptors in given studies
    match_stud   <- match(name.Stud, responseRange[, "study"])
    weight_odors <- responseRange[match_stud, "n_odors"]
    scale        <-
      (sum(weight_receptors * (RMAX / SMAX), na.rm = TRUE) + 
         sum(weight_odors * (RMAX / SMAX), na.rm = TRUE)) / 
      (sum(weight_receptors, na.rm = TRUE) + sum(weight_odors, na.rm = TRUE))
    global_norm   <- scale * MV
    return(global_norm)
  }