Updates to track the original location of duplicated genes

Author: mchaisso

AppendCollapsedDupsList.py

@@ -12,7 +12,7 @@
 
 for d in depth:
     extra=str(max(0,int(d[-1])-2))
-    d[-1] = "-1"
+    d[-1] = "Original"
     d.append(extra)
     d.append("collapse")
     sys.stdout.write("\t".join(d) + "\n")

RemoveOriginal.py

@@ -2,6 +2,9 @@
 import sys
 import re
 inFile = open(sys.argv[1])
+annotate=False
+if len(sys.argv) > 2 and sys.argv[2] == "annotate":
+    annotate=True
 for line in inFile:
     vals=line.split()
     chrom=vals[0]
@@ -12,7 +15,8 @@
     srcChrom=nameRgn[1]
     srcStart=int(nameRgn[2])
     srcEnd=int(nameRgn[3])
+    
     if (chrom == srcChrom and abs(srcStart - start) < 100 and abs(srcEnd-end) < 100):
-        continue
+        sys.stdout.write(line.strip() + "\tORIG\n")
     else:
-        sys.stdout.write(line)
+        sys.stdout.write(line.strip() + "\tCOPY\n")

RemoveOriginalPAF.py

@@ -9,16 +9,15 @@ def Overlap(s1, e1, s2, e2):
     maxS=max(s1,s2)
     minE=min(e1,e2)
     return (minE-maxS)/(e2-s2)
-    
-    
-
-
 
-
+annotate=False
 if sys.argv[1] == "stdin" or sys.argv[1] == "-" or sys.argv[1] == "/sys/stdin":
     inFile = sys.stdin
 else:
     inFile = open(sys.argv[1])
+    if len(sys.argv) > 2:
+        if sys.argv[2] == "annotate":
+            annotate=True
 for line in inFile:
     vals=line.split()
     chrom=vals[5]
@@ -30,7 +29,13 @@ def Overlap(s1, e1, s2, e2):
     srcStart=int(nameRgn[2])
     srcEnd=int(nameRgn[3])
     ovp=Overlap(start, end, srcStart, srcEnd)
+
     if (chrom == srcChrom and ovp > 0.5):
-        continue
+        status="Original"
+        if annotate:            
+            sys.stdout.write(line.strip() + "\tOriginal\n")
     else:
-        sys.stdout.write(line)
+        if annotate:
+            sys.stdout.write(line.strip() + "\tCopy\n")
+        else:
+            sys.stdout.write(line)