R/load_cs_data.R

In atfrank/nmR: Functions for the Analysis and Comparison of NMR Data

 load_cs_data <- function(csfile, remove_assigned=FALSE, accuracyFile=NULL, cs_offset=NULL, residues=NULL, atomBasedWeights=FALSE, names=c("processor","model","resid","resname","nucleus","predCS","expCS","randCS","id")){
	#' Chemical Shift Data Loading Function
	#'
	#' This function allows you to load chemical shift data from file
	#' @param csfile input file name
	#' @param remove_assigned remove data with zero values? Defaults to FALSE
	#' @param accuracyFile expected accuracy file. Defaults to NULL
	#' @param cs_offset file containing offsets. Defaults to NULL
	#' @param residues vector of residues to retain data for. Defaults to NULL
	#' @param atomBasedWeights vector of specifying the residues for to data will be retained. Defaults to NULL
	#' @param names vector of column names. Defaults to c("processor","model","resid","resname","nucleus","predCS","expCS","randCS","id")
	#' @export
	#' @examples
	#' load_cs_data("data/predicted_shifts.txt")

	# read in raw chemical shift data
	cs <- read.table(csfile,col.names=names)

	# remove unassigned chemical shifts
	if(remove_assigned){cs <- subset(cs,expCS!=0)}

	# retain chemical shifts for specified residues
	if (!is.null(residues)) {
			cs <- cs[cs$resid %in% residues,]  
	}

	# retain only protonated chemical shifts
	cs <- subset(cs,!((resname=="ADE")&(nucleus=="C5")))
	cs <- subset(cs,!((resname=="ADE")&(nucleus=="C6")))
	cs <- subset(cs,!((resname=="GUA")&(nucleus=="C5")))
	cs <- subset(cs,!((resname=="GUA")&(nucleus=="C6")))
	cs <- subset(cs,!((resname=="URA")&(nucleus=="C2")))
	cs <- subset(cs,!((resname=="CYT")&(nucleus=="C2")))	
	
	
  # add differential error weights to carbon and proton nuclei
  if(!is.null(accuracyFile)){
		# read in accuracy stats
		accuracyFile <- read.table(accuracyFile, header = TRUE)  
  	cs <- add_cs_weights(cs, accuracyFile, atomBasedWeights)
  }  
	
  # add offets
  if(!is.null(cs_offset)){
		# read in accuracy stats
		cs_offset <- read.table(cs_offset, col.names = c("resname", "nucleus", "offset"))  
		cs <- add_cs_offset(cs, cs_offset) 
  }
	return(cs)
}
 
load_cs_matrix <- function(csfile, skip=1, cs_offset=NULL, residues=NULL, names=c("resid","resname","C1'","C2'","C3'","C4'","C5'","C2","C5","C6","C8","H1'","H2'","H3'","H4'","H2","H5","H5'","H5''","H6","H8","H1","H3","N1","N3")){
   #' Chemical Shift Matrix Loading Function
   #'
   #' This function allows you to load chemical shift data from file in the form of a matrix
   #' @param csfile input file name
   #' @param accuracyFile expected accuracy file. Defaults to NULL
   #' @param cs_offset file containing offsets. Defaults to NULL
   #' @param residues vector of residues to retain data for. Defaults to NULL
   #' @param atomBasedWeights vector of specifying the residues for to data will be retained. Defaults to NULL
   #' @param names vector of column names. Defaults to c("processor","model","resid","resname","nucleus","predCS","expCS","randCS","id")
   #' @export
   #' @examples
   #' load_cs_matrix("data/predicted_shifts.txt")
   d <- read.table(csfile, skip=skip, col.names = names, check.names = FALSE)
   
   cs <- NULL
   # extract chemical shift for each nucleus type
   for (i in 3:length(names)){
     tmp <- d[, c("resname", "resid")]
     tmp$nucleus <-  names[i]
     tmp$predCS <- d[, i]
     tmp$error <- 1
     tmp$weight <- 1    
     if(is.null(data)){
       cs <- tmp
     } else {
       cs <- rbind(cs, tmp)
     }
   }
   
   # retain chemical shifts for specified residues
   if (!is.null(residues)) {
     cs <- cs[cs$resid %in% residues,]  
   }
   
   # retain only protonated chemical shifts
   cs <- subset(cs,!((resname=="ADE")&(nucleus=="C5")))
   cs <- subset(cs,!((resname=="ADE")&(nucleus=="C6")))
   cs <- subset(cs,!((resname=="GUA")&(nucleus=="C5")))
   cs <- subset(cs,!((resname=="GUA")&(nucleus=="C6")))
   cs <- subset(cs,!((resname=="URA")&(nucleus=="C2")))
   cs <- subset(cs,!((resname=="CYT")&(nucleus=="C2")))	
   

   # add offets
   if(!is.null(cs_offset)){
     # read in accuracy stats
     cs_offset <- read.table(cs_offset, col.names = c("resname", "nucleus", "offset"))  
     cs <- add_cs_offset(cs, cs_offset) 
   }
   
   return(cs)
}