diff microsatpurity.py @ 0:70f8259b0b30 draft

Uploaded
author arkarachai-fungtammasan
date Wed, 01 Apr 2015 16:48:58 -0400
parents
children
line wrap: on
line diff
--- /dev/null	Thu Jan 01 00:00:00 1970 +0000
+++ b/microsatpurity.py	Wed Apr 01 16:48:58 2015 -0400
@@ -0,0 +1,24 @@
+import sys
+# remove all read that have impure microsat
+# check only one line at a time
+
+
+fd=open(sys.argv[1])
+lines=fd.xreadlines()
+##motifIx=int(sys.argv[2])
+period=int(sys.argv[2])
+tr_ref_seqIx=int(sys.argv[3])-1
+##output=(sys.argv[4])
+##fout=open(output,'w')
+for line in lines:
+    temp=line.strip().split('\t')
+    temp=filter(None,temp)
+    #motif=temp[motifIx]
+    tr_ref_seq=temp[tr_ref_seqIx]
+    ##period=len(motif)
+    cand_motif=tr_ref_seq[:period]
+    len_microsat=len(tr_ref_seq)
+    expand_microsat_cand=cand_motif*(len_microsat/period) + cand_motif[:(len_microsat%period)]
+    if tr_ref_seq == expand_microsat_cand:
+    	print line.strip()
+        ##print line.strip() >> fout
\ No newline at end of file