Update list-to-extension_config.R

Changes allow extension_config.tsv to list several top terms for the same namespace (there are a few cases).

extension_config-Rscript/list-to-extension_config.R

@@ -7,6 +7,7 @@
     # GO CC - GO:0005575.
 # extension_config.tsv only uses IDs and, therefore, qualifier namespaces must be converted into their topterm IDs, whcih requires screening the FBcv obo file.
 
+rm(list=ls())
 #set work directory as canto-space
 setwd("./")
 getwd()
@@ -101,25 +102,31 @@ colnames(Df1)<-c("domain ID", "subset relation", "extension relation", "range ID
 #read file with list of allowed qualifiers (on "allowed_qualifiers.tsv")
 quallist<-read.table("./extension_config-Rscript/allowed_qualifiers.tsv", sep="\t", header=TRUE, stringsAsFactors=FALSE)
 
-count<-1
 #for each type of allowed qualifier, replace the namespace with corresponding top term with that namespace
+count<-1
+phenpos<-c()
 for(m in 1:nrow(quallist)){
   phen<-unlist(strsplit(quallist[m,3],","))
+  phenqual1<-c()
   for(l in 1:length(phen)){
-    phen[l]<-Df[which(Df[,3]==phen[l]&Df[,5]==phen[l]&is.na(Df[,6])==TRUE),1]
+    phenpos<-which(Df[,3]==phen[l]&Df[,5]==phen[l]&is.na(Df[,6])==TRUE)
+    phenqual<-c()
+    for(l1 in 1:length(phenpos)){
+      phenqual[l1]<-Df[phenpos[l1],1]
+    }
+    phenqual1[l]<-paste(phenqual,collapse = "|")
   }
-  
   if(quallist[m,2]=='phenotype'){
     #if the annotations are for 'phenotypic class' (FBcv:0000347) then:
     # a) the single term 'progressive' shoud be allowed
     if(quallist[m,1]==Df[which(Df[,2]=="phenotypic class"),1]){
       Df1[count,1]<-quallist[m,1]
-      phen<-append(phen, Df[which(Df[,2]=="progressive"), 1])
-      Df1[count,4]<-paste(phen,collapse = '|')
+      phenqual1<-append(phenqual1, Df[which(Df[,2]=="progressive"), 1])
+      Df1[count,4]<-paste(phenqual1,collapse = '|')
       count<-count+1
     # b) and 'the 'fertility/sterility terms must not allow developmental stage terms
     ##once the ontology file is updated so that all viability/mortality/life-span terms are under a common parent term, this next bit should be updated
-      viable<-c("viable","increased mortality","long lived")
+      viable<-c("viable","increased mortality during development","short lived","long lived")
       for (n in 1:length(viable)){ 
       viable[n]<-Df[which(Df[,2]==viable[n]),1]
       }
@@ -138,7 +145,7 @@ for(m in 1:nrow(quallist)){
     }
   else{
     Df1[count,1]<-quallist[m,1]
-    Df1[count,4]<-paste(phen, collapse = '|')
+    Df1[count,4]<-paste(phenqual1, collapse = '|')
     count<-count+1
     Df1[count,1]<-quallist[m,1]
     Df1[count,4]<-"FBdv:00007008"
@@ -150,7 +157,7 @@ for(m in 1:nrow(quallist)){
   }
   if(quallist[m,2]=='genetic interaction'){
     Df1[count,1]<-quallist[m,1]
-    Df1[count,4]<-paste(phen, collapse = '|')
+    Df1[count,4]<-paste(phenqual1, collapse = '|')
     Df1[count,3]<-"int_qual"
     Df1[count,5]<-"Interaction qualifier - ONLY USE FOR INTERACTIONS"
     count<-count+1
@@ -173,4 +180,5 @@ if (is.na(Df1[j,8])==TRUE){Df1[j,8]<-as.character("user")}
   else{}
 }
 
-write.table(Df1, file = "./canto/extension_config.tsv", sep='\t', quote= FALSE, row.names = FALSE)
\ No newline at end of file
+write.table(Df1, file = "./canto/extension_config.tsv", sep='\t', quote= FALSE, row.names = FALSE)
+#for each type of allowed qualifier, replace the namespace with corresponding top term with that namespace