svm validation, tensorflow mp2d pa predictions, r results removed, cleanup

1 files changed, 21 insertions, 7 deletions

diff --git a/scripts/pa-table.rb b/scripts/pa-table.rb
index 8c1037e..1b8ecc8 100755
--- a/scripts/pa-table.rb
+++ b/scripts/pa-table.rb
@@ -2,7 +2,7 @@
 
 # red groups
 tab = []
-File.read("pyrrolizidine-alkaloids/R/PA.RF.outcome.csv").each_line do |l|
+File.read("data/pyrrolizidine-alkaloids/pa-groups.csv").each_line do |l|
   items = l.chomp.split(';')
   if items.first.empty?
     items[0] = "ID"
@@ -17,7 +17,7 @@ end
 tab[0] += ["CID","SMILES","Canonical SMILES","Measured","lazar-MP2D","lazar-MP2D-high-confidence","lazar-CDK","lazar-CDK-high-confidence"]
 
 i = 0
-File.read("pyrrolizidine-alkaloids/180920_PA_complete_SMILES.csv").each_line do |l|
+File.read("data/pyrrolizidine-alkaloids/180920_PA_complete_SMILES.csv").each_line do |l|
   if i > 0
     id,cid,name,smi = l.chomp.split(";")
     tab[i] += [cid,'"'+smi+'"']
@@ -26,7 +26,7 @@ File.read("pyrrolizidine-alkaloids/180920_PA_complete_SMILES.csv").each_line do
 end
 
 i = 0
-File.read("pyrrolizidine-alkaloids/lazar/pa-mp2d-predictions.csv").each_line do |l|
+File.read("pyrrolizidine-alkaloids/mp2d/lazar/pa-mp2d-predictions.csv").each_line do |l|
   if i > 0
     id,cansmi,exp,mut,p0,p1,max_sim,nn = l.chomp.split(",")
     max_sim.to_f < 0.5? hc = "F" : hc = "T"
@@ -36,8 +36,8 @@ File.read("pyrrolizidine-alkaloids/lazar/pa-mp2d-predictions.csv").each_line do
   i += 1
 end
 
-i=0
-File.read("pyrrolizidine-alkaloids/lazar/pa-cdk-predictions.csv").each_line do |l|
+i=1
+File.read("pyrrolizidine-alkaloids/cdk/lazar/pa-cdk-predictions.csv").each_line do |l|
   #if i > 0
     cansmi,exp,mut,p0,p1,max_sim,nn = l.chomp.split(",")
     max_sim.to_f < 0.5? hc = "F" : hc = "T"
@@ -47,6 +47,7 @@ File.read("pyrrolizidine-alkaloids/lazar/pa-cdk-predictions.csv").each_line do |
   i += 1
 end
 
+=begin
 Dir["pyrrolizidine-alkaloids/R/PA.*.outcome.csv"].each do |r|
   tab[0] << "R-"+r.sub('pyrrolizidine-alkaloids/R/PA.','').sub('.outcome.csv','')
   i = 0
@@ -63,9 +64,22 @@ Dir["pyrrolizidine-alkaloids/R/PA.*.outcome.csv"].each do |r|
     i += 1
   end
 end
+=end
 
-Dir["pyrrolizidine-alkaloids/tensorflow/pred.*.v5-ext-Padel-2D.csv"].each do |r|
-  tab[0] << "TF-"+r.sub('pyrrolizidine-alkaloids/tensorflow/pred.','').sub('.v5-ext-Padel-2D.csv','').sub("lr2","LR-scikit").sub("lr","LR-sgd").sub("rf","RF").sub("nn","NN")
+Dir["pyrrolizidine-alkaloids/cdk/tensorflow/*.csv"].each do |r|
+  tab[0] << "TF-"+r.sub('pyrrolizidine-alkaloids/tensorflow/pred.','').sub('.v5-ext-Padel-2D.csv','').sub("lr2","LR-scikit").sub("lr","LR-sgd").sub("rf","RF").sub("nn","NN").sub("svm","SVM")
+  i = 0
+  File.read(r).each_line do |l|
+    if i > 0
+      id,pred = l.chomp.split(",")
+      pred.to_f > 0.5 ? tab[i] << 1 : tab[i] << 0
+    end
+    i += 1
+  end
+end
+
+Dir["pyrrolizidine-alkaloids/cdk/tensorflow/*.csv"].each do |r|
+  tab[0] << "TF-"+r.sub('pyrrolizidine-alkaloids/tensorflow/pred.','').sub('.v5-ext-Padel-2D.csv','').sub("lr2","LR-scikit").sub("lr","LR-sgd").sub("rf","RF").sub("nn","NN").sub("svm","SVM")
   i = 0
   File.read(r).each_line do |l|
     if i > 0