/
command_compute.py
404 lines (321 loc) · 13.9 KB
/
command_compute.py
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
"""
Functions implementing the 'compute' command and related functions.
"""
from __future__ import print_function, division, absolute_import
import os
import os.path
import sys
import random
import screed
import time
from . import sourmash_args
from .signature import SourmashSignature, save_signatures
from .logging import notify, error, set_quiet
from .utils import RustObject
from ._lowlevel import ffi, lib
DEFAULT_COMPUTE_K = '21,31,51'
DEFAULT_LINE_COUNT = 1500
def compute(args):
"""Compute the signature for one or more files.
Use cases:
sourmash compute multiseq.fa => multiseq.fa.sig, etc.
sourmash compute genome.fa --singleton => genome.fa.sig
sourmash compute file1.fa file2.fa -o file.sig
=> creates one output file file.sig, with one signature for each
input file.
sourmash compute file1.fa file2.fa --merge merged -o file.sig
=> creates one output file file.sig, with all sequences from
file1.fa and file2.fa combined into one signature.
"""
set_quiet(args.quiet)
if args.license != 'CC0':
error('error: sourmash only supports CC0-licensed signatures. sorry!')
sys.exit(-1)
if args.input_is_protein and args.dna:
notify('WARNING: input is protein, turning off nucleotide hashing')
args.dna = False
args.protein = True
if args.scaled:
if args.scaled < 1:
error('ERROR: --scaled value must be >= 1')
sys.exit(-1)
if args.scaled != round(args.scaled, 0):
error('ERROR: --scaled value must be integer value')
sys.exit(-1)
if args.scaled >= 1e9:
notify('WARNING: scaled value is nonsensical!? Continuing anyway.')
if args.num_hashes != 0:
notify('setting num_hashes to 0 because --scaled is set')
args.num_hashes = 0
notify('computing signatures for files: {}', ", ".join(args.filenames))
if args.randomize:
notify('randomizing file list because of --randomize')
random.shuffle(args.filenames)
# get list of k-mer sizes for which to compute sketches
ksizes = args.ksizes
notify('Computing signature for ksizes: {}', str(ksizes))
num_sigs = 0
if args.dna and args.protein:
notify('Computing both nucleotide and protein signatures.')
num_sigs = 2*len(ksizes)
elif args.dna and args.dayhoff:
notify('Computing both nucleotide and Dayhoff-encoded protein '
'signatures.')
num_sigs = 2*len(ksizes)
elif args.dna and args.hp:
notify('Computing both nucleotide and Hp-encoded protein '
'signatures.')
num_sigs = 2*len(ksizes)
elif args.dna:
notify('Computing only nucleotide (and not protein) signatures.')
num_sigs = len(ksizes)
elif args.protein:
notify('Computing only protein (and not nucleotide) signatures.')
num_sigs = len(ksizes)
elif args.dayhoff:
notify('Computing only Dayhoff-encoded protein (and not nucleotide) '
'signatures.')
num_sigs = len(ksizes)
elif args.hp:
notify('Computing only hp-encoded protein (and not nucleotide) '
'signatures.')
num_sigs = len(ksizes)
if (args.protein or args.dayhoff or args.hp) and not args.input_is_protein:
bad_ksizes = [ str(k) for k in ksizes if k % 3 != 0 ]
if bad_ksizes:
error('protein ksizes must be divisible by 3, sorry!')
error('bad ksizes: {}', ", ".join(bad_ksizes))
sys.exit(-1)
notify('Computing a total of {} signature(s).', num_sigs)
if num_sigs == 0:
error('...nothing to calculate!? Exiting!')
sys.exit(-1)
if args.merge and not args.output:
error("ERROR: must specify -o with --merge")
sys.exit(-1)
if args.output and args.outdir:
error("ERROR: --outdir doesn't make sense with -o/--output")
sys.exit(-1)
if args.track_abundance:
notify('Tracking abundance of input k-mers.')
if args.merge: # single name specified - combine all
_compute_merged(args)
else: # compute individual signatures
_compute_individual(args)
def _compute_individual(args):
siglist = []
for filename in args.filenames:
sigfile = os.path.basename(filename) + '.sig'
if args.outdir:
sigfile = os.path.join(args.outdir, sigfile)
if not args.output and os.path.exists(sigfile) and not \
args.force:
notify('skipping {} - already done', filename)
continue
if args.singleton:
siglist = []
for n, record in enumerate(screed.open(filename)):
# make minhashes for each sequence
minhashes = make_minhashes(args.ksizes, args.seed, args.protein,
args.dayhoff, args.hp, args.dna,
args.num_hashes,
args.track_abundance, args.scaled)
add_seq(minhashes, record.sequence,
args.input_is_protein, args.check_sequence)
siglist += build_siglist(minhashes, filename, name=record.name)
notify('calculated {} signatures for {} sequences in {}',
len(siglist), n + 1, filename)
elif args.input_is_10x:
from bam2fasta import cli as bam2fasta_cli
# Initializing time
startt = time.time()
metadata = [
"--write-barcode-meta-csv", args.write_barcode_meta_csv] if args.write_barcode_meta_csv else ['', '']
save_fastas = ["--save-fastas", args.save_fastas] if args.save_fastas else ['', '']
barcodes_file = ["--barcodes-file", args.barcodes_file] if args.barcodes_file else ['', '']
rename_10x_barcodes = \
["--rename-10x-barcodes", args.rename_10x_barcodes] if args.rename_10x_barcodes else ['', '']
bam_to_fasta_args = [
'--filename', filename,
'--min-umi-per-barcode', str(args.count_valid_reads),
'--processes', str(args.processes),
'--line-count', str(args.line_count),
barcodes_file[0], barcodes_file[1],
rename_10x_barcodes[0], rename_10x_barcodes[1],
save_fastas[0], save_fastas[1],
metadata[0], metadata[1]]
bam_to_fasta_args = [arg for arg in bam_to_fasta_args if arg != '']
fastas = bam2fasta_cli.convert(bam_to_fasta_args)
# TODO move to bam2fasta since pool imap creates this empty lists and returns them
fastas = [fasta for fasta in fastas if fasta != []]
siglist = []
for fasta in fastas:
for n, record in enumerate(screed.open(fasta)):
# make minhashes for each sequence
minhashes = make_minhashes(args.ksizes, args.seed, args.protein,
args.dayhoff, args.hp, args.dna,
args.num_hashes,
args.track_abundance, args.scaled)
add_seq(minhashes, record.sequence,
args.input_is_protein, args.check_sequence)
siglist += build_siglist(minhashes, fasta, name=record.name)
notify('calculated {} signatures for {} sequences in {}',
len(siglist), n + 1, fasta)
notify("time taken to calculate signature records for 10x file is {:.5f} seconds",
time.time() - startt)
else:
# make minhashes for the whole file
minhashes = make_minhashes(args.ksizes, args.seed, args.protein,
args.dayhoff, args.hp, args.dna,
args.num_hashes,
args.track_abundance, args.scaled)
# consume & calculate signatures
notify('... reading sequences from {}', filename)
name = None
for n, record in enumerate(screed.open(filename)):
if n % 10000 == 0:
if n:
notify('\r...{} {}', filename, n, end='')
elif args.name_from_first:
name = record.name
add_seq(minhashes, record.sequence,
args.input_is_protein, args.check_sequence)
notify('...{} {} sequences', filename, n, end='')
sigs = build_siglist(minhashes, filename, name)
siglist += sigs
notify('calculated {} signatures for {} sequences in {}',
len(sigs), n + 1, filename)
# if no --output specified, save to individual files w/in for loop
if not args.output:
save_siglist(siglist, sigfile)
siglist = []
# if --output specified, all collected signatures => args.output
if args.output:
save_siglist(siglist, args.output)
siglist = []
assert not siglist # juuuust checking.
def _compute_merged(args):
# make minhashes for the whole file
minhashes = make_minhashes(args.ksizes, args.seed, args.protein,
args.dayhoff, args.hp, args.dna,
args.num_hashes,
args.track_abundance, args.scaled)
n = 0
total_seq = 0
for filename in args.filenames:
# consume & calculate signatures
notify('... reading sequences from {}', filename)
for n, record in enumerate(screed.open(filename)):
if n % 10000 == 0 and n:
notify('\r... {} {}', filename, n, end='')
add_seq(minhashes, record.sequence,
args.input_is_protein, args.check_sequence)
notify('... {} {} sequences', filename, n + 1)
total_seq += n + 1
siglist = build_siglist(minhashes, filename, name=args.merge)
notify('calculated {} signatures for {} sequences taken from {} files',
len(siglist), total_seq, len(args.filenames))
# at end, save!
save_siglist(siglist, args.output)
def make_minhashes(ksizes, seed, protein, dayhoff, hp, dna, num_hashes, track_abundance, scaled):
params = ComputeParameters(ksizes, seed, protein, dayhoff, hp, dna, num_hashes, track_abundance, scaled)
sig = SourmashSignature.from_params(params)
return sig
def add_seq(sig, seq, input_is_protein, check_sequence):
if input_is_protein:
sig.add_protein(seq)
else:
sig.add_sequence(seq, not check_sequence)
def build_siglist(sig, filename, name=None):
if name is not None:
sig._name = name
sig.filename = filename
return [sig]
def save_siglist(siglist, sigfile_name):
# save!
with sourmash_args.FileOutput(sigfile_name, 'w') as fp:
save_signatures(siglist, fp)
notify('saved signature(s) to {}. Note: signature license is CC0.',
sigfile_name)
class ComputeParameters(RustObject):
__dealloc_func__ = lib.computeparams_free
def __init__(self, ksizes, seed, protein, dayhoff, hp, dna, num_hashes, track_abundance, scaled):
self._objptr = lib.computeparams_new()
self.seed = seed
self.ksizes = ksizes
self.protein = protein
self.dayhoff = dayhoff
self.hp = hp
self.dna = dna
self.num_hashes = num_hashes
self.track_abundance = track_abundance
self.scaled = scaled
@staticmethod
def from_args(args):
ptr = lib.computeparams_new()
ret = ComputeParameters._from_objptr(ptr)
for arg, value in vars(args).items():
try:
getattr(type(ret), arg).fset(ret, value)
except AttributeError:
pass
return ret
@property
def seed(self):
return self._methodcall(lib.computeparams_seed)
@seed.setter
def seed(self, v):
return self._methodcall(lib.computeparams_set_seed, v)
@property
def ksizes(self):
size = ffi.new("uintptr_t *")
ksizes_ptr = self._methodcall(lib.computeparams_ksizes, size)
size = size[0]
ksizes = ffi.unpack(ksizes_ptr, size)
lib.computeparams_ksizes_free(ksizes_ptr, size)
return ksizes
@ksizes.setter
def ksizes(self, v):
return self._methodcall(lib.computeparams_set_ksizes, list(v), len(v))
@property
def protein(self):
return self._methodcall(lib.computeparams_protein)
@protein.setter
def protein(self, v):
return self._methodcall(lib.computeparams_set_protein, v)
@property
def dayhoff(self):
return self._methodcall(lib.computeparams_dayhoff)
@dayhoff.setter
def dayhoff(self, v):
return self._methodcall(lib.computeparams_set_dayhoff, v)
@property
def hp(self):
return self._methodcall(lib.computeparams_hp)
@hp.setter
def hp(self, v):
return self._methodcall(lib.computeparams_set_hp, v)
@property
def dna(self):
return self._methodcall(lib.computeparams_dna)
@dna.setter
def dna(self, v):
return self._methodcall(lib.computeparams_set_dna, v)
@property
def num_hashes(self):
return self._methodcall(lib.computeparams_num_hashes)
@num_hashes.setter
def num_hashes(self, v):
return self._methodcall(lib.computeparams_set_num_hashes, v)
@property
def track_abundance(self):
return self._methodcall(lib.computeparams_track_abundance)
@track_abundance.setter
def track_abundance(self, v):
return self._methodcall(lib.computeparams_set_track_abundance, v)
@property
def scaled(self):
return self._methodcall(lib.computeparams_scaled)
@scaled.setter
def scaled(self, v):
return self._methodcall(lib.computeparams_set_scaled, int(v))