diff microsatcompat.py @ 0:07588b899c13 draft

Uploaded
author arkarachai-fungtammasan
date Wed, 01 Apr 2015 17:05:51 -0400
parents
children
line wrap: on
line diff
--- /dev/null	Thu Jan 01 00:00:00 1970 +0000
+++ b/microsatcompat.py	Wed Apr 01 17:05:51 2015 -0400
@@ -0,0 +1,36 @@
+import sys
+# remove all read that have unmatch microsat
+# check only one line at a time
+def complement_base(read):
+    collect=''
+    for i in read:
+        if i.upper()=='A':
+            collect+='T'
+        elif i.upper()=='T':
+            collect+='A'
+        elif i.upper()=='C':
+            collect+='G'
+        elif i.upper()=='G':
+            collect+='C'
+    return collect
+   
+def makeallpossible(read):
+    collect=[]
+    for i in range(len(read)):
+        tmp= read[i:]+read[:i]
+        collect.append(tmp)
+        collect.append(complement_base(tmp))
+    return collect
+
+
+fd=open(sys.argv[1])
+lines=fd.xreadlines()
+firstcolumn=int(sys.argv[2])-1 #4
+secondcolumn=int(sys.argv[3])-1 # 10
+for line in lines:
+    temp=line.strip().split('\t')
+    temp=filter(None,temp)
+    micro1=temp[firstcolumn]
+    micro2=temp[secondcolumn]
+    if micro1 in makeallpossible(micro2):
+        print line.strip()
\ No newline at end of file