add average binding filter

1 files changed, 35 insertions, 0 deletions

diff --git a/src/filter_average_binding.py b/src/filter_average_binding.py
new file mode 100644
index 0000000..d250c98
--- /dev/null
+++ b/src/filter_average_binding.py
@@ -0,0 +1,35 @@
+#!/usr/bin/env python
+import sys
+import os
+
+debug = os.environ.get("debug", False)
+
+from subprocess import Popen
+from subprocess import PIPE
+
+def get_length(fn):
+
+	cmd = 'grep "^>" ' + fn + " -v | tr -d '\\n' | wc -c"
+
+	if debug:
+		print "loading sequence end points"
+		print "executing: " + cmd
+
+	points_fh = Popen(cmd, stdout=PIPE, shell=True)
+
+	return int(points_fh.stdout.readline())
+
+if len(sys.argv) < 2:
+	print "filter_average_binding.py foreground.fa binding_distance"
+	exit()
+	
+foreground = sys.argv[1]
+distance = int(sys.argv[2])
+
+genome_length = get_length(foreground)
+
+for line in sys.stdin:
+	(mer, count) = line.split()
+	if (genome_length / int(count)) < distance:
+		sys.stdout.write(line)
+