comparison massdb-helper.R @ 1:45e985cd8e9e draft

planemo upload for repository https://github.com/workflow4metabolomics/lcmsmatching.git commit d4048accde6bdfd5b3e14f5394902d38991854f8-dirty

1:45e985cd8e9e

simplifySpectrum <- function(spec) {
	if(length(spec) == 0){
		return(NA_real_)
	}
	#print(spec)
	if (nrow(spec) == 0)
		return(NA_real_)
	if (ncol(spec) != 2) {
		spec[, BIODB.PEAK.MZ]
		mint <- BIODB.GROUP.INTENSITY %in% colnames(spec)
		pint <- which(mint[1])
		if (length(pint) == 0)
			stop(
				"No intensity column founds, if there is more than 2 column, columns should be named",
				paste0(BIODB.GROUP.INTENSITY, collapse = ", ")
			)
		spec <- spec[, c(BIODB.PEAK.MZ, BIODB.GROUP.INTENSITY[pint[1]])]
		###Normalizing the intenities.
	}
	spec[, 2] <- as.numeric(spec[, 2]) * 100 / max(as.numeric(spec[, 2]))
	colnames(spec) <- c(BIODB.PEAK.MZ, BIODB.PEAK.RELATIVE.INTENSITY)
	spec
}



calcDistance <-
	function(spec1 ,
			 spec2,
			 npmin = 2,
			 fun = c("wcosine"),
			 params = list()) {
		#fun <- match.arg(fun)
		
		#SPec are always notmlized in pourcentage toa voir issues;
		spec1 <- simplifySpectrum(spec1)
		spec2 <- simplifySpectrum(spec2)
		if(is.na(spec1)||is.na(spec2)) return(list(matched=numeric(0),similarity=0))
		params$mz1 <- as.numeric(spec1[, BIODB.PEAK.MZ])
		params$mz2 <- as.numeric(spec2[, BIODB.PEAK.MZ])
		params$int1 <- as.numeric(spec1[, BIODB.PEAK.RELATIVE.INTENSITY])
		params$int2 <- as.numeric(spec2[, BIODB.PEAK.RELATIVE.INTENSITY])
		res <- do.call(fun, args = params)
		if (sum(res$matched != -1) < npmin)
			return(list(matched = res$matched, similarity = 0))
		list(matched = res$matched,
			 similarity = res$measure)
	}



###The returned sim list is not ordered
compareSpectra <-
	function(spec,
			 libspec,
			 npmin = 2,
			 fun = BIODB.MSMS.DIST.WCOSINE,
			 params = list(),
			 decreasing = TRUE) {
		#fun <- match.arg(fun)
		if (length(libspec) == 0) {
			return(NULL)
		}
		if (nrow(spec) == 0) {
			return(NULL)
		}
		
		####spec is directly normalized.
		vall <-
			sapply(
				libspec,
				calcDistance,
				spec1 = spec,
				params = params,
				fun = fun,
				simplify = FALSE
			)
		####the list is ordered with the chosen metric.
		sim <-
			vapply(vall,
				   '[[',
				   i = "similarity",
				   FUN.VALUE = ifelse(decreasing, 0, 1))
		osim <- order(sim, decreasing = decreasing)
		matched <- sapply(vall, '[[', i = "matched", simplify = FALSE)
		
		return(list(
			ord = osim,
			matched = matched,
			similarity = sim
		))
	}