fasta_filter.py
No OneTemporary
Actions

Subscribers

None

File Metadata

	#!/usr/bin/env python3

	from __future__ import print_function
	import sys
	import os
	import re

	if __name__=='__main__':
	helpmess="""Usage:
	fasta_filter seq_list fasta_file [-invert]

	Filters out the sequences in seq_list from fasta_file.
	If -invert is specified, filters out the sequences NOT included.
	"""
	# Inputs
	if len(sys.argv)<3:
	print(helpmess)
	sys.exit(0)
	else:
	infile=os.path.realpath(sys.argv[1])
	fasta=os.path.realpath(sys.argv[2])

	# Output folder
	out_folder='filtered'
	if os.path.exists(out_folder)==False:
	os.system('mkdir %s' % out_folder)

	# Get already sone sequences
	dones = dict()
	with open(infile) as f:
	for line in f:
	val = line.strip()
	dones[val.split()[0]] = val

	# Default name for the file
	basedir,tmpfile=os.path.split(fasta)
	poutname=os.path.join(out_folder, 'inc_'+tmpfile)
	noutname=os.path.join(out_folder, 'mis_'+tmpfile)

	# Copy the content of the file
	with open(poutname, 'w') as p, open(noutname, 'w') as n, open(fasta) as f:
	seq=False
	for line in f:
	if line[0]=='>':
	val = line[1:].strip()
	seq = (val in dones)
	if seq:
	p.write('>{}\n'.format(dones[val]))
	else:
	n.write(line)
	else:
	if seq:
	p.write(line)
	else:
	n.write(line)