updated to allow for multiple drug files

Author: sgosline

build/build_all.py

@@ -91,7 +91,7 @@ def main():
      ### Drug matching scripts take a while
     ### they are their own step and can be run independentyly, before others, or alongside sample/omics
     ### DepMap/Sanger, MPNST, LINCS
-    df=''
+    dflist=[]
     if args.drugs or args.all:
         ###build drug data
         for da in [a for a in datasets if a not in ['cptac','hcmi']]:
@@ -101,8 +101,8 @@ def main():
                 di = da
 
             if not os.path.exists('local/'+da+'_drugs.tsv'):
-                run_cmd([di,'sh','build_drugs.sh',df],da+' drugs')
-            df = '/tmp/'+da+'_drugs.tsv'
+                run_cmd([di,'sh','build_drugs.sh',','.join(dflist)],da+' drugs')
+            dflist = dflist.append('/tmp/'+da+'_drugs.tsv')
 
     #### Any new omics files are created here.
     ## depends on samples!

build/mpnst/02_get_drug_data.R

@@ -19,7 +19,7 @@ if (length(args) == 0) {
 
 # Set your personal access token
 PAT <- args[1]
-olddrugfile <- args[2]
+olddrugfiles <- args[2]
 newdrugfile <- args[3]
 # Log in to Synapse
 synLogin(authToken = PAT)
@@ -71,7 +71,12 @@ alldrugs[which(alldrugs=='PD901')]<-'PD-0325901'
 print(paste(alldrugs,collapse=','))
 
 ##copy old drug to new drug
-file.copy(olddrugfile,newdrugfile)
+olddrugs<-do.call(rbind,lapply(unique(unlist(strsplit(olddrugfiles,split=','))),function(x) read.table(x)))
+olddrugs<-unique(olddrugs)
+
+print(paste('Read in ',nrow(olddrugs),'old drugs'))
+                                        #file.copy(olddrugfile,newdrugfile)
+write.table(olddrugs,file=newdrugfile,sep='\t')
 output_file_path <- newdrugfile
 ignore_file_path <- '/tmp/ignore_chems.txt'