diff HmdbCompound.R @ 0:e66bb061af06 draft

planemo upload for repository https://github.com/workflow4metabolomics/lcmsmatching.git commit 3529b25417f8e1a5836474c9adec4b696d35099d-dirty
author prog
date Tue, 12 Jul 2016 12:02:37 -0400
parents
children 253d531a0193
line wrap: on
line diff
--- /dev/null	Thu Jan 01 00:00:00 1970 +0000
+++ b/HmdbCompound.R	Tue Jul 12 12:02:37 2016 -0400
@@ -0,0 +1,63 @@
+if ( ! exists('HmdbCompound')) { # Do not load again if already loaded
+
+	source('BiodbEntry.R')
+
+	#####################
+	# CLASS DECLARATION #
+	#####################
+
+	HmdbCompound <- setRefClass("HmdbCompound", contains = "BiodbEntry")
+
+	###########
+	# FACTORY #
+	###########
+
+	createHmdbCompoundFromXml <- function(contents, drop = FALSE) {
+
+		library(XML)
+
+		compounds <- list()
+
+		# Define xpath expressions
+		xpath.expr <- character()
+		xpath.expr[[RBIODB.ACCESSION]]          <- "/metabolite/accession"
+		xpath.expr[[RBIODB.KEGG.ID]]            <- "//kegg_id"
+		xpath.expr[[RBIODB.NAME]]               <- "/metabolite/name"
+		xpath.expr[[RBIODB.FORMULA]]            <- "/metabolite/chemical_formula"
+		xpath.expr[[RBIODB.SUPER.CLASS]]        <- "//super_class"
+		xpath.expr[[RBIODB.AVERAGE.MASS]]       <- "//average_molecular_weight"
+		xpath.expr[[RBIODB.MONOISOTOPIC.MASS]]  <- "//monisotopic_moleculate_weight"
+
+		for (content in contents) {
+
+			# Create instance
+			compound <- HmdbCompound$new()
+
+			# Parse XML
+			xml <-  xmlInternalTreeParse(content, asText = TRUE)
+
+			# An error occured
+			if (length(getNodeSet(xml, "//error")) == 0) {
+
+				# Test generic xpath expressions
+				for (field in names(xpath.expr)) {
+					v <- xpathSApply(xml, xpath.expr[[field]], xmlValue)
+					if (length(v) > 0)
+						compound$setField(field, v)
+				}
+
+			}
+
+			compounds <- c(compounds, compound)
+		}
+
+		# Replace elements with no accession id by NULL
+		compounds <- lapply(compounds, function(x) if (is.na(x$getField(RBIODB.ACCESSION))) NULL else x)
+
+		# If the input was a single element, then output a single object
+		if (drop && length(contents) == 1)
+			compounds <- compounds[[1]]
+	
+		return(compounds)
+	}
+}