Mercurial > repos > iuc > khmer_abundance_distribution

annotate filter-below-abund.py @ 1:be929aa4e152 draft

Find changesets by keywords (author, files, the commit message), revision number or hash, or revset expression.
planemo upload for repository https://github.com/galaxyproject/tools-iuc/blob/master/tools/khmer/ commit 72078bced0c08685219c6e80e42ce07458822391
author iuc
date Wed, 11 Nov 2015 10:04:55 -0500
parents c07433531fd3
children 5a2da133262a
Ignore whitespace changes - Everywhere: Within whitespace: At end of lines:
rev   line source
0
c07433531fd3 planemo upload for repository https://github.com/galaxyproject/tools-iuc/blob/master/tools/khmer/ commit be9a20423d1a6ec33d59341e0e61b535127bbce2
iuc
parents:
diff changeset
1 #! /usr/bin/env python
c07433531fd3 planemo upload for repository https://github.com/galaxyproject/tools-iuc/blob/master/tools/khmer/ commit be9a20423d1a6ec33d59341e0e61b535127bbce2
iuc
parents:
diff changeset
2 # This file is part of khmer, https://github.com/dib-lab/khmer/, and is
c07433531fd3 planemo upload for repository https://github.com/galaxyproject/tools-iuc/blob/master/tools/khmer/ commit be9a20423d1a6ec33d59341e0e61b535127bbce2
iuc
parents:
diff changeset
3 # Copyright (C) 2011-2015, Michigan State University.
c07433531fd3 planemo upload for repository https://github.com/galaxyproject/tools-iuc/blob/master/tools/khmer/ commit be9a20423d1a6ec33d59341e0e61b535127bbce2
iuc
parents:
diff changeset
4 # Copyright (C) 2015, The Regents of the University of California.
c07433531fd3 planemo upload for repository https://github.com/galaxyproject/tools-iuc/blob/master/tools/khmer/ commit be9a20423d1a6ec33d59341e0e61b535127bbce2
iuc
parents:
diff changeset
5 #
c07433531fd3 planemo upload for repository https://github.com/galaxyproject/tools-iuc/blob/master/tools/khmer/ commit be9a20423d1a6ec33d59341e0e61b535127bbce2
iuc
parents:
diff changeset
6 # Redistribution and use in source and binary forms, with or without
c07433531fd3 planemo upload for repository https://github.com/galaxyproject/tools-iuc/blob/master/tools/khmer/ commit be9a20423d1a6ec33d59341e0e61b535127bbce2
iuc
parents:
diff changeset
7 # modification, are permitted provided that the following conditions are
c07433531fd3 planemo upload for repository https://github.com/galaxyproject/tools-iuc/blob/master/tools/khmer/ commit be9a20423d1a6ec33d59341e0e61b535127bbce2
iuc
parents:
diff changeset
8 # met:
c07433531fd3 planemo upload for repository https://github.com/galaxyproject/tools-iuc/blob/master/tools/khmer/ commit be9a20423d1a6ec33d59341e0e61b535127bbce2
iuc
parents:
diff changeset
9 #
c07433531fd3 planemo upload for repository https://github.com/galaxyproject/tools-iuc/blob/master/tools/khmer/ commit be9a20423d1a6ec33d59341e0e61b535127bbce2
iuc
parents:
diff changeset
10 # * Redistributions of source code must retain the above copyright
c07433531fd3 planemo upload for repository https://github.com/galaxyproject/tools-iuc/blob/master/tools/khmer/ commit be9a20423d1a6ec33d59341e0e61b535127bbce2
iuc
parents:
diff changeset
11 # notice, this list of conditions and the following disclaimer.
c07433531fd3 planemo upload for repository https://github.com/galaxyproject/tools-iuc/blob/master/tools/khmer/ commit be9a20423d1a6ec33d59341e0e61b535127bbce2
iuc
parents:
diff changeset
12 #
c07433531fd3 planemo upload for repository https://github.com/galaxyproject/tools-iuc/blob/master/tools/khmer/ commit be9a20423d1a6ec33d59341e0e61b535127bbce2
iuc
parents:
diff changeset
13 # * Redistributions in binary form must reproduce the above
c07433531fd3 planemo upload for repository https://github.com/galaxyproject/tools-iuc/blob/master/tools/khmer/ commit be9a20423d1a6ec33d59341e0e61b535127bbce2
iuc
parents:
diff changeset
14 # copyright notice, this list of conditions and the following
c07433531fd3 planemo upload for repository https://github.com/galaxyproject/tools-iuc/blob/master/tools/khmer/ commit be9a20423d1a6ec33d59341e0e61b535127bbce2
iuc
parents:
diff changeset
15 # disclaimer in the documentation and/or other materials provided
c07433531fd3 planemo upload for repository https://github.com/galaxyproject/tools-iuc/blob/master/tools/khmer/ commit be9a20423d1a6ec33d59341e0e61b535127bbce2
iuc
parents:
diff changeset
16 # with the distribution.
c07433531fd3 planemo upload for repository https://github.com/galaxyproject/tools-iuc/blob/master/tools/khmer/ commit be9a20423d1a6ec33d59341e0e61b535127bbce2
iuc
parents:
diff changeset
17 #
c07433531fd3 planemo upload for repository https://github.com/galaxyproject/tools-iuc/blob/master/tools/khmer/ commit be9a20423d1a6ec33d59341e0e61b535127bbce2
iuc
parents:
diff changeset
18 # * Neither the name of the Michigan State University nor the names
c07433531fd3 planemo upload for repository https://github.com/galaxyproject/tools-iuc/blob/master/tools/khmer/ commit be9a20423d1a6ec33d59341e0e61b535127bbce2
iuc
parents:
diff changeset
19 # of its contributors may be used to endorse or promote products
c07433531fd3 planemo upload for repository https://github.com/galaxyproject/tools-iuc/blob/master/tools/khmer/ commit be9a20423d1a6ec33d59341e0e61b535127bbce2
iuc
parents:
diff changeset
20 # derived from this software without specific prior written
c07433531fd3 planemo upload for repository https://github.com/galaxyproject/tools-iuc/blob/master/tools/khmer/ commit be9a20423d1a6ec33d59341e0e61b535127bbce2
iuc
parents:
diff changeset
21 # permission.
c07433531fd3 planemo upload for repository https://github.com/galaxyproject/tools-iuc/blob/master/tools/khmer/ commit be9a20423d1a6ec33d59341e0e61b535127bbce2
iuc
parents:
diff changeset
22 #
c07433531fd3 planemo upload for repository https://github.com/galaxyproject/tools-iuc/blob/master/tools/khmer/ commit be9a20423d1a6ec33d59341e0e61b535127bbce2
iuc
parents:
diff changeset
23 # THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
c07433531fd3 planemo upload for repository https://github.com/galaxyproject/tools-iuc/blob/master/tools/khmer/ commit be9a20423d1a6ec33d59341e0e61b535127bbce2
iuc
parents:
diff changeset
24 # "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
c07433531fd3 planemo upload for repository https://github.com/galaxyproject/tools-iuc/blob/master/tools/khmer/ commit be9a20423d1a6ec33d59341e0e61b535127bbce2
iuc
parents:
diff changeset
25 # LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
c07433531fd3 planemo upload for repository https://github.com/galaxyproject/tools-iuc/blob/master/tools/khmer/ commit be9a20423d1a6ec33d59341e0e61b535127bbce2
iuc
parents:
diff changeset
26 # A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
c07433531fd3 planemo upload for repository https://github.com/galaxyproject/tools-iuc/blob/master/tools/khmer/ commit be9a20423d1a6ec33d59341e0e61b535127bbce2
iuc
parents:
diff changeset
27 # HOLDER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
c07433531fd3 planemo upload for repository https://github.com/galaxyproject/tools-iuc/blob/master/tools/khmer/ commit be9a20423d1a6ec33d59341e0e61b535127bbce2
iuc
parents:
diff changeset
28 # SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
c07433531fd3 planemo upload for repository https://github.com/galaxyproject/tools-iuc/blob/master/tools/khmer/ commit be9a20423d1a6ec33d59341e0e61b535127bbce2
iuc
parents:
diff changeset
29 # LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
c07433531fd3 planemo upload for repository https://github.com/galaxyproject/tools-iuc/blob/master/tools/khmer/ commit be9a20423d1a6ec33d59341e0e61b535127bbce2
iuc
parents:
diff changeset
30 # DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
c07433531fd3 planemo upload for repository https://github.com/galaxyproject/tools-iuc/blob/master/tools/khmer/ commit be9a20423d1a6ec33d59341e0e61b535127bbce2
iuc
parents:
diff changeset
31 # THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
c07433531fd3 planemo upload for repository https://github.com/galaxyproject/tools-iuc/blob/master/tools/khmer/ commit be9a20423d1a6ec33d59341e0e61b535127bbce2
iuc
parents:
diff changeset
32 # (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
c07433531fd3 planemo upload for repository https://github.com/galaxyproject/tools-iuc/blob/master/tools/khmer/ commit be9a20423d1a6ec33d59341e0e61b535127bbce2
iuc
parents:
diff changeset
33 # OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
c07433531fd3 planemo upload for repository https://github.com/galaxyproject/tools-iuc/blob/master/tools/khmer/ commit be9a20423d1a6ec33d59341e0e61b535127bbce2
iuc
parents:
diff changeset
34 #
c07433531fd3 planemo upload for repository https://github.com/galaxyproject/tools-iuc/blob/master/tools/khmer/ commit be9a20423d1a6ec33d59341e0e61b535127bbce2
iuc
parents:
diff changeset
35 # Contact: khmer-project@idyll.org
c07433531fd3 planemo upload for repository https://github.com/galaxyproject/tools-iuc/blob/master/tools/khmer/ commit be9a20423d1a6ec33d59341e0e61b535127bbce2
iuc
parents:
diff changeset
36 from __future__ import print_function
c07433531fd3 planemo upload for repository https://github.com/galaxyproject/tools-iuc/blob/master/tools/khmer/ commit be9a20423d1a6ec33d59341e0e61b535127bbce2
iuc
parents:
diff changeset
37 import sys
c07433531fd3 planemo upload for repository https://github.com/galaxyproject/tools-iuc/blob/master/tools/khmer/ commit be9a20423d1a6ec33d59341e0e61b535127bbce2
iuc
parents:
diff changeset
38 import os
c07433531fd3 planemo upload for repository https://github.com/galaxyproject/tools-iuc/blob/master/tools/khmer/ commit be9a20423d1a6ec33d59341e0e61b535127bbce2
iuc
parents:
diff changeset
39 import khmer
c07433531fd3 planemo upload for repository https://github.com/galaxyproject/tools-iuc/blob/master/tools/khmer/ commit be9a20423d1a6ec33d59341e0e61b535127bbce2
iuc
parents:
diff changeset
40 from khmer.thread_utils import ThreadedSequenceProcessor, verbose_fasta_iter
c07433531fd3 planemo upload for repository https://github.com/galaxyproject/tools-iuc/blob/master/tools/khmer/ commit be9a20423d1a6ec33d59341e0e61b535127bbce2
iuc
parents:
diff changeset
41
c07433531fd3 planemo upload for repository https://github.com/galaxyproject/tools-iuc/blob/master/tools/khmer/ commit be9a20423d1a6ec33d59341e0e61b535127bbce2
iuc
parents:
diff changeset
42 WORKER_THREADS = 8
c07433531fd3 planemo upload for repository https://github.com/galaxyproject/tools-iuc/blob/master/tools/khmer/ commit be9a20423d1a6ec33d59341e0e61b535127bbce2
iuc
parents:
diff changeset
43 GROUPSIZE = 100
c07433531fd3 planemo upload for repository https://github.com/galaxyproject/tools-iuc/blob/master/tools/khmer/ commit be9a20423d1a6ec33d59341e0e61b535127bbce2
iuc
parents:
diff changeset
44
c07433531fd3 planemo upload for repository https://github.com/galaxyproject/tools-iuc/blob/master/tools/khmer/ commit be9a20423d1a6ec33d59341e0e61b535127bbce2
iuc
parents:
diff changeset
45 CUTOFF = 50
c07433531fd3 planemo upload for repository https://github.com/galaxyproject/tools-iuc/blob/master/tools/khmer/ commit be9a20423d1a6ec33d59341e0e61b535127bbce2
iuc
parents:
diff changeset
46
c07433531fd3 planemo upload for repository https://github.com/galaxyproject/tools-iuc/blob/master/tools/khmer/ commit be9a20423d1a6ec33d59341e0e61b535127bbce2
iuc
parents:
diff changeset
47 ###
c07433531fd3 planemo upload for repository https://github.com/galaxyproject/tools-iuc/blob/master/tools/khmer/ commit be9a20423d1a6ec33d59341e0e61b535127bbce2
iuc
parents:
diff changeset
48
c07433531fd3 planemo upload for repository https://github.com/galaxyproject/tools-iuc/blob/master/tools/khmer/ commit be9a20423d1a6ec33d59341e0e61b535127bbce2
iuc
parents:
diff changeset
49
c07433531fd3 planemo upload for repository https://github.com/galaxyproject/tools-iuc/blob/master/tools/khmer/ commit be9a20423d1a6ec33d59341e0e61b535127bbce2
iuc
parents:
diff changeset
50 def main():
c07433531fd3 planemo upload for repository https://github.com/galaxyproject/tools-iuc/blob/master/tools/khmer/ commit be9a20423d1a6ec33d59341e0e61b535127bbce2
iuc
parents:
diff changeset
51 counting_ht = sys.argv[1]
c07433531fd3 planemo upload for repository https://github.com/galaxyproject/tools-iuc/blob/master/tools/khmer/ commit be9a20423d1a6ec33d59341e0e61b535127bbce2
iuc
parents:
diff changeset
52 infiles = sys.argv[2:]
c07433531fd3 planemo upload for repository https://github.com/galaxyproject/tools-iuc/blob/master/tools/khmer/ commit be9a20423d1a6ec33d59341e0e61b535127bbce2
iuc
parents:
diff changeset
53
c07433531fd3 planemo upload for repository https://github.com/galaxyproject/tools-iuc/blob/master/tools/khmer/ commit be9a20423d1a6ec33d59341e0e61b535127bbce2
iuc
parents:
diff changeset
54 print('file with ht: %s' % counting_ht)
c07433531fd3 planemo upload for repository https://github.com/galaxyproject/tools-iuc/blob/master/tools/khmer/ commit be9a20423d1a6ec33d59341e0e61b535127bbce2
iuc
parents:
diff changeset
55 print('-- settings:')
c07433531fd3 planemo upload for repository https://github.com/galaxyproject/tools-iuc/blob/master/tools/khmer/ commit be9a20423d1a6ec33d59341e0e61b535127bbce2
iuc
parents:
diff changeset
56 print('N THREADS', WORKER_THREADS)
c07433531fd3 planemo upload for repository https://github.com/galaxyproject/tools-iuc/blob/master/tools/khmer/ commit be9a20423d1a6ec33d59341e0e61b535127bbce2
iuc
parents:
diff changeset
57 print('--')
c07433531fd3 planemo upload for repository https://github.com/galaxyproject/tools-iuc/blob/master/tools/khmer/ commit be9a20423d1a6ec33d59341e0e61b535127bbce2
iuc
parents:
diff changeset
58
c07433531fd3 planemo upload for repository https://github.com/galaxyproject/tools-iuc/blob/master/tools/khmer/ commit be9a20423d1a6ec33d59341e0e61b535127bbce2
iuc
parents:
diff changeset
59 print('making hashtable')
c07433531fd3 planemo upload for repository https://github.com/galaxyproject/tools-iuc/blob/master/tools/khmer/ commit be9a20423d1a6ec33d59341e0e61b535127bbce2
iuc
parents:
diff changeset
60 ht = khmer.load_countgraph(counting_ht)
c07433531fd3 planemo upload for repository https://github.com/galaxyproject/tools-iuc/blob/master/tools/khmer/ commit be9a20423d1a6ec33d59341e0e61b535127bbce2
iuc
parents:
diff changeset
61 K = ht.ksize()
c07433531fd3 planemo upload for repository https://github.com/galaxyproject/tools-iuc/blob/master/tools/khmer/ commit be9a20423d1a6ec33d59341e0e61b535127bbce2
iuc
parents:
diff changeset
62
c07433531fd3 planemo upload for repository https://github.com/galaxyproject/tools-iuc/blob/master/tools/khmer/ commit be9a20423d1a6ec33d59341e0e61b535127bbce2
iuc
parents:
diff changeset
63 for infile in infiles:
c07433531fd3 planemo upload for repository https://github.com/galaxyproject/tools-iuc/blob/master/tools/khmer/ commit be9a20423d1a6ec33d59341e0e61b535127bbce2
iuc
parents:
diff changeset
64 print('filtering', infile)
c07433531fd3 planemo upload for repository https://github.com/galaxyproject/tools-iuc/blob/master/tools/khmer/ commit be9a20423d1a6ec33d59341e0e61b535127bbce2
iuc
parents:
diff changeset
65 outfile = os.path.basename(infile) + '.below'
c07433531fd3 planemo upload for repository https://github.com/galaxyproject/tools-iuc/blob/master/tools/khmer/ commit be9a20423d1a6ec33d59341e0e61b535127bbce2
iuc
parents:
diff changeset
66
c07433531fd3 planemo upload for repository https://github.com/galaxyproject/tools-iuc/blob/master/tools/khmer/ commit be9a20423d1a6ec33d59341e0e61b535127bbce2
iuc
parents:
diff changeset
67 outfp = open(outfile, 'w')
c07433531fd3 planemo upload for repository https://github.com/galaxyproject/tools-iuc/blob/master/tools/khmer/ commit be9a20423d1a6ec33d59341e0e61b535127bbce2
iuc
parents:
diff changeset
68
c07433531fd3 planemo upload for repository https://github.com/galaxyproject/tools-iuc/blob/master/tools/khmer/ commit be9a20423d1a6ec33d59341e0e61b535127bbce2
iuc
parents:
diff changeset
69 def process_fn(record, ht=ht):
c07433531fd3 planemo upload for repository https://github.com/galaxyproject/tools-iuc/blob/master/tools/khmer/ commit be9a20423d1a6ec33d59341e0e61b535127bbce2
iuc
parents:
diff changeset
70 name = record['name']
c07433531fd3 planemo upload for repository https://github.com/galaxyproject/tools-iuc/blob/master/tools/khmer/ commit be9a20423d1a6ec33d59341e0e61b535127bbce2
iuc
parents:
diff changeset
71 seq = record['sequence']
c07433531fd3 planemo upload for repository https://github.com/galaxyproject/tools-iuc/blob/master/tools/khmer/ commit be9a20423d1a6ec33d59341e0e61b535127bbce2
iuc
parents:
diff changeset
72 if 'N' in seq:
c07433531fd3 planemo upload for repository https://github.com/galaxyproject/tools-iuc/blob/master/tools/khmer/ commit be9a20423d1a6ec33d59341e0e61b535127bbce2
iuc
parents:
diff changeset
73 return None, None
c07433531fd3 planemo upload for repository https://github.com/galaxyproject/tools-iuc/blob/master/tools/khmer/ commit be9a20423d1a6ec33d59341e0e61b535127bbce2
iuc
parents:
diff changeset
74
c07433531fd3 planemo upload for repository https://github.com/galaxyproject/tools-iuc/blob/master/tools/khmer/ commit be9a20423d1a6ec33d59341e0e61b535127bbce2
iuc
parents:
diff changeset
75 trim_seq, trim_at = ht.trim_below_abundance(seq, CUTOFF)
c07433531fd3 planemo upload for repository https://github.com/galaxyproject/tools-iuc/blob/master/tools/khmer/ commit be9a20423d1a6ec33d59341e0e61b535127bbce2
iuc
parents:
diff changeset
76
c07433531fd3 planemo upload for repository https://github.com/galaxyproject/tools-iuc/blob/master/tools/khmer/ commit be9a20423d1a6ec33d59341e0e61b535127bbce2
iuc
parents:
diff changeset
77 if trim_at >= K:
c07433531fd3 planemo upload for repository https://github.com/galaxyproject/tools-iuc/blob/master/tools/khmer/ commit be9a20423d1a6ec33d59341e0e61b535127bbce2
iuc
parents:
diff changeset
78 return name, trim_seq
c07433531fd3 planemo upload for repository https://github.com/galaxyproject/tools-iuc/blob/master/tools/khmer/ commit be9a20423d1a6ec33d59341e0e61b535127bbce2
iuc
parents:
diff changeset
79
c07433531fd3 planemo upload for repository https://github.com/galaxyproject/tools-iuc/blob/master/tools/khmer/ commit be9a20423d1a6ec33d59341e0e61b535127bbce2
iuc
parents:
diff changeset
80 return None, None
c07433531fd3 planemo upload for repository https://github.com/galaxyproject/tools-iuc/blob/master/tools/khmer/ commit be9a20423d1a6ec33d59341e0e61b535127bbce2
iuc
parents:
diff changeset
81
c07433531fd3 planemo upload for repository https://github.com/galaxyproject/tools-iuc/blob/master/tools/khmer/ commit be9a20423d1a6ec33d59341e0e61b535127bbce2
iuc
parents:
diff changeset
82 tsp = ThreadedSequenceProcessor(process_fn, WORKER_THREADS, GROUPSIZE)
c07433531fd3 planemo upload for repository https://github.com/galaxyproject/tools-iuc/blob/master/tools/khmer/ commit be9a20423d1a6ec33d59341e0e61b535127bbce2
iuc
parents:
diff changeset
83
c07433531fd3 planemo upload for repository https://github.com/galaxyproject/tools-iuc/blob/master/tools/khmer/ commit be9a20423d1a6ec33d59341e0e61b535127bbce2
iuc
parents:
diff changeset
84 tsp.start(verbose_fasta_iter(infile), outfp)
c07433531fd3 planemo upload for repository https://github.com/galaxyproject/tools-iuc/blob/master/tools/khmer/ commit be9a20423d1a6ec33d59341e0e61b535127bbce2
iuc
parents:
diff changeset
85
c07433531fd3 planemo upload for repository https://github.com/galaxyproject/tools-iuc/blob/master/tools/khmer/ commit be9a20423d1a6ec33d59341e0e61b535127bbce2
iuc
parents:
diff changeset
86 if __name__ == '__main__':
c07433531fd3 planemo upload for repository https://github.com/galaxyproject/tools-iuc/blob/master/tools/khmer/ commit be9a20423d1a6ec33d59341e0e61b535127bbce2
iuc
parents:
diff changeset
87 main()