lib.py 32.4 KB
Newer Older
jhoogenboom's avatar
jhoogenboom committed
1
2
#!/usr/bin/env python

jhoogenboom's avatar
jhoogenboom committed
3
import re, sys, argparse
jhoogenboom's avatar
jhoogenboom committed
4
#import numpy as np  # Imported only when calling nnls()
jhoogenboom's avatar
jhoogenboom committed
5
6

from ConfigParser import RawConfigParser, MissingSectionHeaderError
7
from StringIO import StringIO
jhoogenboom's avatar
jhoogenboom committed
8
from collections import defaultdict
jhoogenboom's avatar
jhoogenboom committed
9
10
11
12

# Patterns that match entire sequences.
PAT_SEQ_RAW = re.compile("^[ACGT]*$")
PAT_SEQ_TSSV = re.compile("^(?:[ACGT]+\(\d+\))*$")
13
14
PAT_SEQ_ALLELENAME = re.compile(  # First line: n_ACT[m] or alias.
    "^(?:(?:(?:CE)?\d+(?:\.\d+)?_(?:[ACGT]+\[\d+\])+)|((?!_).+?))"
jhoogenboom's avatar
jhoogenboom committed
15
16
17
    "(?:_[-+]\d+(?:\.1)?(?P<a>(?:(?<=\.1)-)|(?<!\.1)[ACGT]+)>"  # _+3A>
        "(?!(?P=a))(?:[ACTG]+|-))*$")  # Portion of variants after '>'.

18
# Patterns that match blocks of TSSV-style sequences and allele names.
jhoogenboom's avatar
jhoogenboom committed
19
20
PAT_TSSV_BLOCK = re.compile("([ACGT]+)\((\d+)\)")
PAT_ALLELENAME_BLOCK = re.compile("([ACGT]+)\[(\d+)\]")
21
PAT_ALIAS = re.compile("^(?!_).+$")
jhoogenboom's avatar
jhoogenboom committed
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36

# Pattern that matches a single prefix/suffix variant.
PAT_VARIANT = re.compile(
    "^([-+]\d+)(?:\.1)?"  # Position number
    "(?P<a>(?:(?<=\.1)-)|(?<!\.1)[ACGT]+)>"  # From part
    "(?!(?P=a))([ACTG]+|-)$")  # To part

# Patterns that match (parts of) an STR definition.
PAT_STR_DEF = re.compile(
    "^(?:[ACGT]+\s+\d+\s+\d+(?:\s+[ACGT]+\s+\d+\s+\d+)*)?$")
PAT_STR_DEF_BLOCK = re.compile("([ACGT]+)\s+(\d+)\s+(\d+)")

# Pattern to split a comma-, semicolon-, or space-separated list.
PAT_SPLIT = re.compile("[,; ]+")

jhoogenboom's avatar
jhoogenboom committed
37
38
39
40
41
42
43
44
# Default regular expression to capture sample tags in file names.
# This is the default of the -e command line option.
DEF_TAG_EXPR = "^(.+?)(?:\.[^.]+)?$"

# Default formatting template to write sample tags.
# This is the default of the -f command line option.
DEF_TAG_FORMAT = "\\1"

jhoogenboom's avatar
jhoogenboom committed
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

def call_variants(template, sequence, reverse_indices=False, cache=True,
                  debug=False):
    """
    Perform a global alignment of sequence to template and return a
    list of variants detected.  All variants are given as substitutions
    in the form posX>Y, where the first base in the template is pos=1.
    Set reverse_indices to True to count from right to left instead.
    Insertions and deletions are pos.1->Y and posX>-, respectively.

    By default, the results of this function are cached.  Set cache to
    False to suppress caching the result and reduce memory usage.

    Setting debug to True will cause the alignment matrices to be
    printed to sys.stdout.  Be aware that they can be quite large.
    """
    # Saving the results in a cache to avoid repeating alignments.
    try:
        return call_variants.cache[template, sequence, reverse_indices]
    except KeyError:
        pass

    row_offset = len(template) + 1
    matrix_match = [0] * row_offset * (len(sequence)+1)
    matrix_gap1 = [-sys.maxint-1] * row_offset * (len(sequence)+1)
    matrix_gap2 = [-sys.maxint-1] * row_offset * (len(sequence)+1)

    MATCH_SCORE = 1
    MISMATCH_SCORE = -1
    GAP_OPEN_SCORE = -10
    GAP_EXTEND_SCORE = -1

    for i in range(len(matrix_match)):
        x = i % row_offset
        y = i / row_offset

        # Initialisation of first row and column.
        if x == 0 or y == 0:
            if x != 0:
                # Top row.
                matrix_gap1[i] = GAP_OPEN_SCORE + GAP_EXTEND_SCORE * (x - 1)
                matrix_match[i] = matrix_gap1[i]
            elif y != 0:
                # Left column.
                matrix_gap2[i] = GAP_OPEN_SCORE + GAP_EXTEND_SCORE * (y - 1)
                matrix_match[i] = matrix_gap2[i]
            continue

        matrix_gap1[i] = max(
            matrix_match[i-1] + GAP_OPEN_SCORE,
            matrix_gap1[i-1] + GAP_EXTEND_SCORE)
        matrix_gap2[i] = max(
            matrix_match[i-row_offset] + GAP_OPEN_SCORE,
            matrix_gap2[i-row_offset] + GAP_EXTEND_SCORE)

        if template[x-1] == sequence[y-1]:
            match = MATCH_SCORE
        else:
            match = MISMATCH_SCORE

        matrix_match[i] = max(
            matrix_match[i-1-row_offset] + match,
            matrix_gap1[i],
            matrix_gap2[i])

    if debug:
        print("GAP1")
        for i in range(0, len(matrix_gap1), row_offset):
            print(("%5i" * row_offset) % tuple(matrix_gap1[i:i+row_offset]))
        print("GAP2")
        for i in range(0, len(matrix_gap2), row_offset):
            print(("%5i" * row_offset) % tuple(matrix_gap2[i:i+row_offset]))
        print("Match")
        for i in range(0, len(matrix_match), row_offset):
            print(("%5i" * row_offset) % tuple(matrix_match[i:i+row_offset]))


    # Backtracking.
    variants = []
    variant_template = 0
    variant_sequence = 0
    i = len(matrix_match) - 1
    while i >= 0:
        x = i % row_offset
        y = i / row_offset

        if matrix_gap1[i] == matrix_match[i]:
            # Go horizontally.  Deletion.
            variant_template += 1
            i -= 1
            continue

        if matrix_gap2[i] == matrix_match[i]:
            # Go vertically.  Insertion.
            variant_sequence += 1
            i -= row_offset
            continue

        # Only backtracking diagonally if a gap is out of the question.
        # Go diagonally.  Either match or mismatch.
        if i == 0 or template[x - 1] == sequence[y - 1]:
            # Match.  Flush variants.
            if variant_template or variant_sequence:
                if variant_template == 0:
                    # Insertions: "-131.1->C" instead of "-130->C".
                    variants.append("%+i.1->%s" % (
                        x - int(reverse_indices) * row_offset,
                        sequence[y:y+variant_sequence]))
                else:
                    variants.append("%+i%s>%s" % (
                        (x + 1) - int(reverse_indices) * row_offset,
                        template[x:x+variant_template],
                        sequence[y:y+variant_sequence] or "-"))
                variant_template = 0
                variant_sequence = 0
        else:
            # Start/extend mismatch.
            variant_template += 1
            variant_sequence += 1
        i -= 1 + row_offset

    # If reverse_indices=False, we need to reverse the output instead.
    if not reverse_indices:
        variants.reverse()

    # Store the result in the cache.
    if cache:
        call_variants.cache[template, sequence, reverse_indices] = variants
    return variants
#call_variants
call_variants.cache = {}


def mutate_sequence(sequence, variants):
    """Apply the given variants to the given sequence."""
    if not sequence and len(variants) > 1:
        raise ValueError("With an empty sequence, only a single variant is "
                         "possible: an insertion '+0.1->x' or '-1.1->x'.")
    sequence = list(sequence)
    for variant in variants:
        vm = PAT_VARIANT.match(variant)
        if vm is None:
            raise ValueError("Unrecognised variant '%s'" % variant)
        pos = int(vm.group(1))
        old = vm.group(2)
        new = vm.group(3)
        if old == "-":
            old = ""
        if new == "-":
            new = ""
        if pos < 0:
            pos += len(sequence) + 1
        if pos == 0 and len(sequence) == 0:
            # Insertion into empty sequence.
            return new
        if old != "".join(sequence[pos-1:pos+len(old)-1]):
            raise ValueError("Incorrect original sequence in variant '%s'; "
                             "should be '%s'!" % (variant,
                               "".join(sequence[pos-1:pos+len(old)-1]) or "-"))
        sequence[pos-1:pos+len(old)-1] = [""] * len(old)
        if pos:
            sequence[pos-1] += new
        else:
            # Insertion at the beginning of the sequence
            sequence[0] = new + sequence[0]
    return "".join(sequence)
#mutate_sequence


def parse_library(handle):
215
216
217
    if handle == sys.stdin:
        # Can't seek on pipes, so read it into a buffer first.
        handle = StringIO(sys.stdin.read())
jhoogenboom's avatar
jhoogenboom committed
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
    try:
        return parse_library_ini(handle)
    except MissingSectionHeaderError:
        # Not an ini file.
        pass
    handle.seek(0)
    return parse_library_tsv(handle)
#parse_library


def parse_library_tsv(handle):
    """
    Parse a TSSV library file (tab-separated values format).

    The provided file should contain at least four columns: marker name,
    left flanking sequence, right flanking sequence, and STR definition.

    Return a nested dict with top-level keys "flanks" and "regex".
    """
    library = {
      "flanks": {},
239
240
      "regex": {},
      "regex_middle": {}
jhoogenboom's avatar
jhoogenboom committed
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
    }
    for line in handle:
        line = map(lambda x: x.strip(), line.rstrip("\r\n").split("\t"))
        if line == [""]:
            continue
        if len(line) < 4:
            raise ValueError(
                "Invalid library file: encountered line with %i columns, "
                "need at least 4" % len(line))
        marker = line[0]
        if PAT_SEQ_RAW.match(line[1]) is None:
            raise ValueError("Flanking sequence '%s' of marker %s is invalid" %
                             (line[1], marker))
        if PAT_SEQ_RAW.match(line[2]) is None:
            raise ValueError("Flanking sequence '%s' of marker %s is invalid" %
                             (line[2], marker))
        if PAT_STR_DEF.match(line[3]) is None:
            raise ValueError("STR definition '%s' of marker %s is invalid" %
                             (line[3], marker))
        library["flanks"][marker] = line[1:3]
261
262
263
264
        library["regex_middle"][marker] = re.compile("".join(map(lambda x:
            "(%s){%s,%s}" % x, PAT_STR_DEF_BLOCK.findall(line[3]))))
        library["regex"][marker] = re.compile(
            "".join(["^", library["regex_middle"][marker].pattern, "$"]))
jhoogenboom's avatar
jhoogenboom committed
265
266
267
268
269
270
271
272
273
274
    return library
#parse_library_tsv


def parse_library_ini(handle):
    library = {
      "flanks": {},
      "prefix": {},
      "suffix": {},
      "regex": {},
275
      "regex_middle": {},
jhoogenboom's avatar
jhoogenboom committed
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
      "length_adjust": defaultdict(int),
      "block_length": defaultdict(lambda: 4),
      "aliases": {}
    }
    markers = set()

    ini = RawConfigParser()
    ini.optionxform = str
    ini.readfp(handle)
    for section in ini.sections():
        for marker in ini.options(section):
            value = ini.get(section, marker)
            if section == "flanks":
                values = PAT_SPLIT.split(value)
                if len(values) != 2:
                    raise ValueError(
                        "For marker %s, %i flanking sequences were given,"
                        "need exactly 2" % (marker, len(values)))
                for value in values:
                    if PAT_SEQ_RAW.match(value) is None:
                        raise ValueError(
                            "Flanking sequence '%s' of marker %s is invalid" %
                            (value, marker))
                library["flanks"][marker] = values
                markers.add(marker)
            elif section == "prefix":
                values = PAT_SPLIT.split(value)
                for value in values:
                    if PAT_SEQ_RAW.match(value) is None:
                        raise ValueError(
                            "Prefix sequence '%s' of marker %s is invalid" %
                            (value, marker))
308
                library["prefix"][marker] = values
jhoogenboom's avatar
jhoogenboom committed
309
310
311
312
313
314
315
316
                markers.add(marker)
            elif section == "suffix":
                values = PAT_SPLIT.split(value)
                for value in values:
                    if PAT_SEQ_RAW.match(value) is None:
                        raise ValueError(
                            "Suffix sequence '%s' of marker %s is invalid" %
                            (value, marker))
317
                library["suffix"][marker] = values
jhoogenboom's avatar
jhoogenboom committed
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
                markers.add(marker)
            elif section == "length_adjust":
                try:
                    value = int(value)
                except:
                    raise ValueError(
                        "Length adjustment '%s' of marker %s is not a valid "
                        "integer" % (value, marker))
                library["length_adjust"][marker] = value
                markers.add(marker)
            elif section == "block_length":
                try:
                    value = int(value)
                except:
                    raise ValueError(
                        "Block length '%s' of marker %s is not a valid integer"
                        % (value, marker))
                library["block_length"][marker] = value
                markers.add(marker)
            elif section == "aliases":
                values = PAT_SPLIT.split(value)
                if len(values) != 3:
                    raise ValueError("Alias %s does not have 3 values, but %i"
                                     % (marker, len(values)))
                if PAT_SEQ_RAW.match(values[1]) is None:
343
344
345
346
347
348
349
                    raise ValueError(
                        "Alias sequence '%s' of alias %s is invalid" %
                        (values[1], marker))
                if PAT_ALIAS.match(values[2]) is None:
                    raise ValueError(
                        "Allele name '%s' of alias %s is invalid" %
                        (values[2], marker))
jhoogenboom's avatar
jhoogenboom committed
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
                library["aliases"][marker] = {
                    "marker": values[0],
                    "sequence": values[1],
                    "name": values[2]
                }
                markers.add(marker)
            elif section == "repeat":
                if PAT_STR_DEF.match(value) is None:
                    raise ValueError(
                        "STR definition '%s' of marker %s is invalid" %
                        (value, marker))
                library["regex"][marker] = value
                markers.add(marker)

    # Compile regular expressions.
    # NOTE: The libconvert tool expects "(seq){num,num}" blocks ONLY!
    # TODO: Should a single prefix/suffix be required (i.e., seq{1,1})?
    #       Then also update libconvert when converting to TSSV format.
    for marker in markers:
        parts = []
370
        partsm = []
jhoogenboom's avatar
jhoogenboom committed
371
372
373
374
        if marker in library["prefix"]:
            parts += map(lambda x: "(%s){0,1}" % x, library["prefix"][marker])
        if marker in library["aliases"]:
            parts.append("(%s){0,1}" % library["aliases"][marker]["sequence"])
375
            partsm.append("(%s){0,1}" % library["aliases"][marker]["sequence"])
jhoogenboom's avatar
jhoogenboom committed
376
        elif marker in library["regex"]:
377
            partsm = map(lambda x: "(%s){%s,%s}" % x,
jhoogenboom's avatar
jhoogenboom committed
378
                         PAT_STR_DEF_BLOCK.findall(library["regex"][marker]))
379
            parts += partsm
jhoogenboom's avatar
jhoogenboom committed
380
381
382
383
384
        if marker in library["suffix"]:
            parts += map(lambda x: "(%s){0,1}" % x, library["suffix"][marker])
        if parts:
            library["regex"][marker] = re.compile(
                "".join(["^"] + parts + ["$"]))
385
386
        if partsm:
            library["regex_middle"][marker] = re.compile("".join(partsm))
jhoogenboom's avatar
jhoogenboom committed
387
388
    return library
#parse_library_ini
jhoogenboom's avatar
jhoogenboom committed
389
390


391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
def load_profiles(profilefile, library=None):
    profiles = {}

    # Read the profile file without assuming it is sorted.
    for line in profilefile:
        line = line.rstrip("\r\n").split("\t")
        if line == [""]:
            continue
        if len(line) < 3:
            raise ValueError(
                "Invalid background noise profiles file: encountered line "
                "with %i columns, need at least 3" % len(line))
        marker = line[0]
        try:
            num = int(line[1])
        except ValueError:
            raise ValueError(
                "Invalid background noise profiles file: encountered "
                "non-number '%s' in the second column" % line[1])
        values = line[2:]
        if marker not in profiles:
            profiles[marker] = {
                "m": len(values),
                "n": abs(num),
                "seqs": [],
                "forward": {},
                "reverse": {}
                }
        elif len(values) != profiles[marker]["m"]:
            raise ValueError(
                "Invalid background noise profiles file: profiles of "
                "marker '%s'% have inconsistent lengths" % marker)
        profiles[marker]["n"] = max(abs(num), profiles[marker]["n"])
        if num == 0:
            if profiles[marker]["seqs"]:
                raise ValueError(
                    "Invalid background profiles noise file: encountered "
                    "multiple header lines for marker '%s'" % marker)
            values = map(
                lambda seq: ensure_sequence_format(
                    seq, "raw", library=library, marker=marker),
                values)
            profiles[marker]["seqs"] = values
        else:
            direction = "forward" if num > 0 else "reverse"
            if abs(num) in profiles[marker][direction]:
                raise ValueError(
                    "Invalid background profiles noise file: encountered "
                    "multiple %s profiles for marker '%s' allele %i" %
                    (direction, marker, abs(num)))
            values = map(float, values)
            profiles[marker][direction][abs(num)] = values

    # Check completeness and reorder true alleles.
    for marker in profiles:
        newprofiles = {"forward": [], "reverse": []}
        if not profiles[marker]["seqs"]:
            raise ValueError(
                "Invalid background profiles noise file: missing header line "
                "for marker '%s'" % marker)
        for i in range(1, profiles[marker]["n"] + 1):
            for direction in newprofiles:
                if i not in profiles[marker][direction]:
                    raise ValueError(
                        "Invalid background profiles noise file: missing %s "
                        "profile for marker '%s' allele %i" %
                        (direction, marker, i))
                newprofiles[direction].append(profiles[marker][direction][i])
        profiles[marker]["forward"] = newprofiles["forward"]
        profiles[marker]["reverse"] = newprofiles["reverse"]

    return profiles
#load_profiles


466
467
468
469
470
471
472
473
def regex_longest_match(pattern, subject):
    """Return the longest match of the pattern in the subject string."""
    return reduce(lambda x, y:
        y if x is None or x.end()-x.start() < y.end()-y.start() else x,
        pattern.finditer(subject), None)
#regex_longest_match


jhoogenboom's avatar
jhoogenboom committed
474
475
476
def detect_sequence_format(seq):
    """Return format of seq.  One of 'raw', 'tssv', or 'allelename'."""
    if not seq:
jhoogenboom's avatar
jhoogenboom committed
477
        raise ValueError("Empty sequence")
jhoogenboom's avatar
jhoogenboom committed
478
479
480
481
482
483
    if PAT_SEQ_RAW.match(seq):
        return 'raw'
    if PAT_SEQ_TSSV.match(seq):
        return 'tssv'
    if PAT_SEQ_ALLELENAME.match(seq):
        return 'allelename'
jhoogenboom's avatar
jhoogenboom committed
484
    raise ValueError("Unrecognised sequence format")
jhoogenboom's avatar
jhoogenboom committed
485
486
487
#detect_sequence_format


jhoogenboom's avatar
jhoogenboom committed
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
def convert_sequence_tssv_raw(seq):
    return "".join(map(
        lambda block: block[0] * int(block[1]),
        PAT_TSSV_BLOCK.findall(seq)))
#convert_sequence_tssv_raw


def convert_sequence_raw_tssv(seq, library, marker, return_alias=False):
    # Try to match this marker's pattern, or any of its aliases.
    match = None
    if "aliases" in library:
        for alias in library["aliases"]:
            if (library["aliases"][alias]["marker"] == marker and
                    alias in library["regex"]):
                match = library["regex"][alias].match(seq)
                if match is not None:
                    marker = alias
                    break
    if match is None and marker in library["regex"]:
        match = library["regex"][marker].match(seq)
508
509
510
    if match is not None:
        parts = [(match.group(i), match.end(i))
                 for i in range(1, match.lastindex+1) if match.group(i)]
jhoogenboom's avatar
jhoogenboom committed
511

512
    # Use heuristics if the sequence does not match the pattern.
jhoogenboom's avatar
jhoogenboom committed
513
    else:
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558

        # Find explictily provided prefix and/or suffix if present.
        pre_suf = ["", ""]
        if "prefix" in library and marker in library["prefix"]:
            for prefix in library["prefix"][marker]:
                if seq.startswith(prefix):
                    pre_suf[0] = prefix
                    seq = seq[len(prefix):]
                    break
        if "suffix" in library and marker in library["suffix"]:
            for suffix in library["suffix"][marker]:
                if seq.endswith(suffix):
                    pre_suf[1] = suffix
                    seq = seq[:-len(suffix)]
                    break

        # Find longest match of middle pattern.
        middle = [(seq, len(pre_suf[0])+len(seq))]
        if marker in library["regex_middle"]:
            match = regex_longest_match(library["regex_middle"][marker], seq)
            if match is not None and match.end()-match.start():
                # TODO: If I did not have a prefix yet, but the
                # canonical prefix ends with the first few blocks that
                # we obtained in the match, move these blocks into the
                # prefix.  Also do this with the suffix.
                middle = [(match.group(i), match.end(i)+len(pre_suf[0]))
                    for i in range(1, match.lastindex+1) if match.group(i)]
                pre_suf[0] += seq[:match.start()]
                pre_suf[1] = seq[match.end():] + pre_suf[1]
                seq = match.group()

        # Now construct parts.
        parts = []
        if pre_suf[0]:
            parts.append((pre_suf[0], len(pre_suf[0])))
        parts += middle
        if pre_suf[1]:
            parts.append((pre_suf[1], sum(map(len,pre_suf))+len(seq)))

    seq = reduce(
        lambda a, b: (a[0] + "%s(%i)" % (b[0], (b[1]-a[1])/len(b[0])), b[1]),
        reduce(
            lambda x, y: x[:-1] + [y] if x[-1][0] == y[0] else x + [y],
            parts,
            [("", 0)]))[0]
jhoogenboom's avatar
jhoogenboom committed
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
    return (seq, marker) if return_alias else seq
#convert_sequence_raw_tssv


def convert_sequence_allelename_tssv(seq, library, marker):
    # Check whether there is an alias for this sequence.
    if "aliases" in library:
        for alias in library["aliases"]:
            if library["aliases"][alias]["marker"] == marker and (
                    seq == library["aliases"][alias]["name"] or
                    seq.startswith(library["aliases"][alias]["name"] + "_")):
                marker = alias
                seq = "".join([
                    "0_",
                    library["aliases"][alias]["sequence"] + "[1]",
                    seq[len(library["aliases"][alias]["name"]):]])
                break

577
578
579
580
581
    # It should NOT be an alias now.
    match = PAT_SEQ_ALLELENAME.match(seq)
    if match is None or match.group(1) is not None:
        raise ValueError("Invalid allele name '%s' encountered!" % seq)

jhoogenboom's avatar
jhoogenboom committed
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
    allele = seq.split("_")

    # Get and mutate prefix and suffix.
    if "prefix" in library and marker in library["prefix"]:
        prefix = library["prefix"][marker][0]
    else:
        prefix = ""
    if "suffix" in library and marker in library["suffix"]:
        suffix = library["suffix"][marker][0]
    else:
        suffix = ""
    variants = [[], []]
    for variant in allele[2:]:
        if variant[0] == "-":
            if not prefix:
                raise ValueError("Encountered prefix variant '%s', but marker "
                                 "'%s' has no prefix!" % (variant, marker))
            variants[0].append(variant)
        elif variant[0] == "+":
            if not suffix:
                raise ValueError("Encountered suffix variant '%s', but marker "
                                 "'%s' has no suffix!" % (variant, marker))
            variants[1].append(variant)
        else:
            raise ValueError("Unrecognised variant '%s'" % variant)
    if variants[0]:
        prefix = mutate_sequence(prefix, variants[0])
    if variants[1]:
        suffix = mutate_sequence(suffix, variants[1])

    blocks = []
    if prefix:
        blocks.append((prefix, 1))
    for block in PAT_ALLELENAME_BLOCK.findall(allele[1]):
        blocks.append((block[0], int(block[1])))
    if suffix:
        blocks.append((suffix, 1))
    return "".join(map(lambda block: "%s(%i)" % block, blocks))
#convert_sequence_allelename_tssv


def convert_sequence_raw_allelename(seq, library, marker):
    # We actually convert raw->allelename via TSSV format.
    seq, alias = convert_sequence_raw_tssv(seq, library, marker, True)
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
    blocks = PAT_TSSV_BLOCK.findall(seq)

    # Find prefix and suffix.
    prefix = suffix = this_prefix = this_suffix = ""
    remaining_blocks = len(blocks)
    if "prefix" in library and marker in library["prefix"]:
        prefix = library["prefix"][marker][0]
        remaining_blocks -= 1
    if "suffix" in library and marker in library["suffix"]:
        suffix = library["suffix"][marker][0]
        remaining_blocks -= 1
    if remaining_blocks > 0 and prefix and blocks[0][1] == "1":
        this_prefix = blocks[0][0]
        blocks = blocks[1:]
    if remaining_blocks > 0 and suffix and blocks[-1][1] == "1":
        this_suffix = blocks[-1][0]
        blocks = blocks[:-1]
jhoogenboom's avatar
jhoogenboom committed
643
644
645
646

    # Generate prefix/suffix variants.
    length = 0
    variants = []
647
648
649
650
651
652
    if prefix != this_prefix:
        variants += call_variants(prefix, this_prefix, True)
        length += len(this_prefix) - len(prefix)
    if suffix != this_suffix:
        variants += call_variants(suffix, this_suffix, False)
        length += len(this_suffix) - len(suffix)
jhoogenboom's avatar
jhoogenboom committed
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713
714
715
716
717
718

    # We are ready to return the allele name of aliases.
    if alias != marker:
        return "_".join([library["aliases"][marker]["name"]] + variants)

    # Compute CE allele number for the other alleles.
    # TODO: perhaps produce a more intelligent name if there is exactly
    #       one alias with the same length
    blocknames = []
    if "block_length" in library:
        blocksize = library["block_length"][marker]
    else:
        blocksize = 4
    if "length_adjust" in library:
        length -= library["length_adjust"][marker]
    for block in blocks:
        blocknames.append("%s[%s]" % (block[0], block[1]))
        length += len(block[0]) * int(block[1])

    allelename = "CE" + str(length / blocksize)
    if length % blocksize:
        allelename += "." + str(length % blocksize)
    return "_".join([allelename, "".join(blocknames)] + variants)
#convert_sequence_raw_allelename


def ensure_sequence_format(seq, to_format, from_format=None, library=None,
                           marker=None):
    """Convert seq to 'raw', 'tssv', or 'allelename' format."""
    known_formats = ["raw", "tssv", "allelename"]
    if to_format not in known_formats:
        raise ValueError("Unknown format '%s', choose from %s" %
                         (to_format, known_formats))
    if from_format is None:
        from_format = detect_sequence_format(seq) if seq else "raw"
    elif from_format not in known_formats:
        raise ValueError("Unknown format '%s', choose from %s" %
                         (from_format, known_formats))

    # No conversion needed?
    if to_format == from_format:
        return seq

    # From TSSV to raw sequence is easy.
    # We'll need a library and marker name for anything else.
    if (library is None or marker is None) and (from_format != "tssv" or
            to_format != "raw"):
        raise ValueError("Sequence needs to be converted from %s to %s, this "
                          "conversion requires a library file" %
                          (from_format, to_format))

    # Perform conversions.
    if from_format == "allelename":
        seq = convert_sequence_allelename_tssv(seq, library, marker)
    if to_format == "tssv":
        if from_format == "raw":
            return convert_sequence_raw_tssv(seq, library, marker)
        return seq
    if from_format != "raw":
        seq = convert_sequence_tssv_raw(seq)
    if to_format == "raw":
        return seq
    return convert_sequence_raw_allelename(seq, library, marker)
#ensure_sequence_format


jhoogenboom's avatar
jhoogenboom committed
719
720
721
722
723
724
725
726
727
728
729
730
731
732
733
734
735
736
737
738
739
740
741
742
743
744
745
746
747
748
749
750
751
752
753
754
755
756
757
758
759
760
761
762
763
def nnls(A, C, B=None, max_iter=200, min_change=0.0001, debug=False):
    """
    Solve for B in A * B = C in the least squares sense, s.t. B >= 0.

    Hint: call nnls(B.T, C.T).T to solve for A.

    Algorithm has converged if the sum of squared error has decreased
    by less than a factor of min_change in one iteration.  If debug is
    True, print the sum of squared error to sys.stdout after each
    iteration.

    This code was partially adopted from nimfa.methods.factorization.bd.
    """
    import numpy as np
    if B is None:
        B = np.matrix(np.zeros([A.shape[1], C.shape[1]]))
    E = A.T * A
    F = A.T * C
    prev_score = cur_score = sys.float_info.max
    for i in range(max_iter):
        for n in range(B.shape[0]):
            nn = list(range(n)) + list(range(n + 1, B.shape[0]))
            tmp = (F[n, :] - E[n, nn] * B[nn, :]) / E[n, n]
            tmp[np.isnan(tmp)] = 0
            tmp[tmp < 0] = 0
            B[n, :] = tmp
        prev_score = cur_score
        cur_score = np.square(C - A * B).sum()
        score_change = (prev_score-cur_score)/prev_score

        if debug:
            if i:
                print("%4i %15.6f %15.6f %6.2f" % (i, cur_score,
                    prev_score-cur_score, 100*score_change))
            else:
                print("%4i %15.6f" % (i, cur_score))

        if not cur_score or score_change < min_change:
            # We have converged.
            break

    return B
#nnls


jhoogenboom's avatar
jhoogenboom committed
764
765
766
767
768
769
770
771
772
773
774
775
776
777
def get_column_ids(column_names, *names):
    """Find all names in column_names and return their indices."""
    result = []
    for name in names:
        try:
            result.append(column_names.index(name))
        except ValueError:
            raise Exception("Column not found in input file: %s" % name)
    if len(result) == 1:
        return result[0]
    return tuple(result)
#get_column_ids


jhoogenboom's avatar
jhoogenboom committed
778
779
780
781
782
783
784
785
786
787
788
789
790
791
792
793
794
795
796
797
798
799
800
def parse_allelelist(allelelist, convert=None, library=None):
    """Read allele list from open file handle."""
    column_names = allelelist.readline().rstrip("\r\n").split("\t")
    colid_sample, colid_marker, colid_allele = get_column_ids(column_names,
        "sample", "marker", "allele")
    alleles = {}
    for line in allelelist:
        line = line.rstrip("\r\n").split("\t")
        sample = line[colid_sample]
        marker = line[colid_marker]
        allele = line[colid_allele]
        if convert is not None:
            allele = ensure_sequence_format(allele, convert, library=library,
                                            marker=marker)
        if sample not in alleles:
            alleles[sample] = {}
        if marker not in alleles[sample]:
            alleles[sample][marker] = set()
        alleles[sample][marker].add(allele)
    return alleles
#parse_allelelist


jhoogenboom's avatar
jhoogenboom committed
801
802
803
804
805
806
807
808
809
def pos_int_arg(value):
    """Convert str to int, raise ArgumentTypeError if not positive."""
    if not value.isdigit() or not int(value):
        raise argparse.ArgumentTypeError(
            "invalid positive int value: '%s'" % value)
    return int(value)
#pos_int_arg


jhoogenboom's avatar
jhoogenboom committed
810
811
812
813
814
815
816
817
818
819
820
821
822
823
824
825
826
827
828
829
830
831
832
833
834
835
836
837
838
839
840
841
842
843
844
845
846
847
848
849
850
851
852
853
854
855
856
857
858
859
860
def add_allele_detection_args(parser):
    parser.add_argument('-a', '--allelelist', metavar="ALLELEFILE",
        type=argparse.FileType('r'),
        help="file containing a list of the true alleles of each sample")
    parser.add_argument('-c', '--annotation-column', metavar="COLNAME",
        help="name of a column in the sample files, which contains a value "
             "beginning with 'ALLELE' for the true alleles of the sample")
#add_allele_detection_args


def add_sample_files_args(parser):
    """Add arguments for opening sample files to the given parser."""
    parser.add_argument('filelist', nargs='*', metavar="FILE",
        default=[sys.stdin], type=argparse.FileType('r'),
        help="the data file(s) to process (default: read from stdin)")
    parser.add_argument('-e', '--tag-expr', metavar="REGEX", type=re.compile,
        default=DEF_TAG_EXPR,
        help="regular expression that captures (using one or more capturing "
             "groups) the sample tags from the file names; by default, the "
             "entire file name except for its extension (if any) is captured")
    parser.add_argument('-f', '--tag-format', metavar="EXPR",
        default=DEF_TAG_FORMAT,
        help="format of the sample tags produced; a capturing group reference "
             "like '\\n' refers to the n-th capturing group in the regular "
             "expression specified with -e/--tag-expr (the default of '\\1' "
             "simply uses the first capturing group); with a single sample, "
             "you can enter the samle tag here explicitly")
#add_sample_fils_args


def get_tag(filename, tag_expr, tag_format):
    """Return formatted sample tag from filename using regex."""
    try:
        return tag_expr.search(filename).expand(tag_format)
    except:
        return filename
#get_tag


def map_tags_to_files(filelist, tag_expr, tag_format):
    tags_to_files = {}
    for infile in filelist:
        tag = get_tag(infile.name, tag_expr, tag_format)
        if tag not in tags_to_files:
            tags_to_files[tag] = [infile]
        else:
            tags_to_files[tag].append(infile)
    return tags_to_files
#map_tags_to_files


jhoogenboom's avatar
jhoogenboom committed
861
862
863
864
865
def print_db(text, debug):
    """Print text if debug is True."""
    if debug:
        print(text)
#print_db