scripts/last-dotplot
author Martin C. Frith
Thu Mar 02 10:10:51 2017 +0900 (2017-03-02)
changeset 839 bbc6f00e683b
parent 838 7b972338fe04
child 840 85a72978fb7d
permissions -rwxr-xr-x
Added trim options to last-dotplot.
Martin@1
     1
#! /usr/bin/env python
Martin@1
     2
Martin@272
     3
# Read pair-wise alignments in MAF or LAST tabular format: write an
Martin@272
     4
# "Oxford grid", a.k.a. dotplot.
Martin@1
     5
Martin@1
     6
# TODO: Currently, pixels with zero aligned nt-pairs are white, and
Martin@1
     7
# pixels with one or more aligned nt-pairs are black.  This can look
Martin@1
     8
# too crowded for large genome alignments.  I tried shading each pixel
Martin@1
     9
# according to the number of aligned nt-pairs within it, but the
Martin@1
    10
# result is too faint.  How can this be done better?
Martin@1
    11
Martin@651
    12
import fileinput, fnmatch, itertools, optparse, os, re, sys
Martin@475
    13
Martin@475
    14
# Try to make PIL/PILLOW work:
Martin@475
    15
try: from PIL import Image, ImageDraw, ImageFont, ImageColor
Martin@475
    16
except ImportError: import Image, ImageDraw, ImageFont, ImageColor
Martin@1
    17
Martin@644
    18
def warn(message):
Martin@644
    19
    prog = os.path.basename(sys.argv[0])
Martin@644
    20
    sys.stderr.write(prog + ": " + message + "\n")
Martin@644
    21
Martin@482
    22
def tabBlocks(beg1, beg2, blocks):
Martin@482
    23
    '''Get the gapless blocks of an alignment, from LAST tabular format.'''
Martin@482
    24
    for i in blocks.split(","):
Martin@482
    25
        if ":" in i:
Martin@482
    26
            x, y = i.split(":")
Martin@482
    27
            beg1 += int(x)
Martin@482
    28
            beg2 += int(y)
Martin@482
    29
        else:
Martin@482
    30
            size = int(i)
Martin@482
    31
            yield beg1, beg2, size
Martin@482
    32
            beg1 += size
Martin@482
    33
            beg2 += size
Martin@272
    34
Martin@482
    35
def mafBlocks(beg1, beg2, seq1, seq2):
Martin@482
    36
    '''Get the gapless blocks of an alignment, from MAF format.'''
Martin@482
    37
    size = 0
Martin@482
    38
    for x, y in itertools.izip(seq1, seq2):
Martin@482
    39
        if x == "-":
Martin@482
    40
            if size:
Martin@482
    41
                yield beg1, beg2, size
Martin@482
    42
                beg1 += size
Martin@482
    43
                beg2 += size
Martin@482
    44
                size = 0
Martin@482
    45
            beg2 += 1
Martin@482
    46
        elif y == "-":
Martin@482
    47
            if size:
Martin@482
    48
                yield beg1, beg2, size
Martin@482
    49
                beg1 += size
Martin@482
    50
                beg2 += size
Martin@482
    51
                size = 0
Martin@482
    52
            beg1 += 1
Martin@272
    53
        else:
Martin@482
    54
            size += 1
Martin@482
    55
    if size: yield beg1, beg2, size
Martin@272
    56
Martin@482
    57
def alignmentInput(lines):
Martin@482
    58
    '''Get alignments and sequence lengths, from MAF or tabular format.'''
Martin@482
    59
    mafCount = 0
Martin@272
    60
    for line in lines:
Martin@272
    61
        w = line.split()
Martin@272
    62
        if line[0].isdigit():  # tabular format
Martin@482
    63
            chr1, beg1, seqlen1 = w[1], int(w[2]), int(w[5])
Martin@482
    64
            if w[4] == "-": beg1 -= seqlen1
Martin@482
    65
            chr2, beg2, seqlen2 = w[6], int(w[7]), int(w[10])
Martin@482
    66
            if w[9] == "-": beg2 -= seqlen2
Martin@839
    67
            blocks = list(tabBlocks(beg1, beg2, w[11]))
Martin@482
    68
            yield chr1, seqlen1, chr2, seqlen2, blocks
Martin@272
    69
        elif line[0] == "s":  # MAF format
Martin@482
    70
            if mafCount == 0:
Martin@482
    71
                chr1, beg1, seqlen1, seq1 = w[1], int(w[2]), int(w[5]), w[6]
Martin@482
    72
                if w[4] == "-": beg1 -= seqlen1
Martin@482
    73
                mafCount = 1
Martin@482
    74
            else:
Martin@482
    75
                chr2, beg2, seqlen2, seq2 = w[1], int(w[2]), int(w[5]), w[6]
Martin@482
    76
                if w[4] == "-": beg2 -= seqlen2
Martin@839
    77
                blocks = list(mafBlocks(beg1, beg2, seq1, seq2))
Martin@482
    78
                yield chr1, seqlen1, chr2, seqlen2, blocks
Martin@482
    79
                mafCount = 0
Martin@272
    80
Martin@651
    81
def isWantedSequenceName(name, patterns):
Martin@651
    82
    if not patterns: return True
Martin@651
    83
    base = name.split(".")[-1]  # allow for names like hg19.chr7
Martin@651
    84
    for i in patterns:
Martin@651
    85
        if fnmatch.fnmatchcase(name, i): return True
Martin@651
    86
        if fnmatch.fnmatchcase(base, i): return True
Martin@651
    87
    return False
Martin@651
    88
Martin@839
    89
def updateSeqLimits(isTrim, seqLimits, seqName, seqLen, blocks, index):
Martin@839
    90
    if isTrim:
Martin@839
    91
        beg = blocks[0][index]
Martin@839
    92
        end = blocks[-1][index] + blocks[-1][2]
Martin@839
    93
        if beg < 0: beg, end = -end, -beg
Martin@839
    94
        if seqName in seqLimits:
Martin@839
    95
            b, e = seqLimits[seqName]
Martin@839
    96
            seqLimits[seqName] = min(b, beg), max(e, end)
Martin@839
    97
        else:
Martin@839
    98
            seqLimits[seqName] = beg, end
Martin@839
    99
    else:
Martin@839
   100
        seqLimits[seqName] = 0, seqLen
Martin@839
   101
Martin@651
   102
def readAlignments(fileName, opts):
Martin@839
   103
    '''Get alignments and sequence limits, from MAF or tabular format.'''
Martin@482
   104
    alignments = []
Martin@839
   105
    seqLimits1 = {}
Martin@839
   106
    seqLimits2 = {}
Martin@651
   107
    lines = fileinput.input(fileName)
Martin@838
   108
    for seqName1, seqLen1, seqName2, seqLen2, blocks in alignmentInput(lines):
Martin@838
   109
        if not isWantedSequenceName(seqName1, opts.seq1): continue
Martin@838
   110
        if not isWantedSequenceName(seqName2, opts.seq2): continue
Martin@838
   111
        aln = seqName1, seqName2, blocks
Martin@482
   112
        alignments.append(aln)
Martin@839
   113
        updateSeqLimits(opts.trim1, seqLimits1, seqName1, seqLen1, blocks, 0)
Martin@839
   114
        updateSeqLimits(opts.trim2, seqLimits2, seqName2, seqLen2, blocks, 1)
Martin@839
   115
    return alignments, seqLimits1, seqLimits2
Martin@1
   116
Martin@1
   117
def natural_sort_key(my_string):
Martin@1
   118
    '''Return a sort key for "natural" ordering, e.g. chr9 < chr10.'''
Martin@1
   119
    parts = re.split(r'(\d+)', my_string)
Martin@1
   120
    parts[1::2] = map(int, parts[1::2])
Martin@1
   121
    return parts
Martin@1
   122
Martin@647
   123
def get_text_sizes(my_strings, font, fontsize, image_mode):
Martin@1
   124
    '''Get widths & heights, in pixels, of some strings.'''
Martin@647
   125
    if fontsize == 0: return [(0, 0) for i in my_strings]
Martin@1
   126
    image_size = 1, 1
Martin@134
   127
    im = Image.new(image_mode, image_size)
Martin@1
   128
    draw = ImageDraw.Draw(im)
Martin@1
   129
    return [draw.textsize(i, font=font) for i in my_strings]
Martin@1
   130
Martin@839
   131
def get_seq_info(seqLimits, font, fontsize, image_mode):
Martin@1
   132
    '''Return miscellaneous information about the sequences.'''
Martin@839
   133
    seqNames = seqLimits.keys()
Martin@837
   134
    seqNames.sort(key=natural_sort_key)
Martin@839
   135
    seq_sizes = [seqLimits[i][1] - seqLimits[i][0] for i in seqNames]
Martin@837
   136
    name_sizes = get_text_sizes(seqNames, font, fontsize, image_mode)
Martin@28
   137
    margin = max(zip(*name_sizes)[1])  # maximum text height
Martin@837
   138
    return seqNames, seq_sizes, name_sizes, margin
Martin@1
   139
Martin@1
   140
def div_ceil(x, y):
Martin@1
   141
    '''Return x / y rounded up.'''
Martin@1
   142
    q, r = divmod(x, y)
Martin@1
   143
    return q + (r != 0)
Martin@1
   144
Martin@1
   145
def tot_seq_pix(seq_sizes, bp_per_pix):
Martin@1
   146
    '''Return the total pixels needed for sequences of the given sizes.'''
Martin@28
   147
    return sum([div_ceil(i, bp_per_pix) for i in seq_sizes])
Martin@1
   148
Martin@645
   149
def get_bp_per_pix(seq_sizes, pix_tween_seqs, pix_limit):
Martin@1
   150
    '''Get the minimum bp-per-pixel that fits in the size limit.'''
Martin@1
   151
    seq_num = len(seq_sizes)
Martin@1
   152
    seq_pix_limit = pix_limit - pix_tween_seqs * (seq_num - 1)
Martin@1
   153
    if seq_pix_limit < seq_num:
Martin@649
   154
        raise Exception("can't fit the image: too many sequences?")
Martin@51
   155
    lower_bound = div_ceil(sum(seq_sizes), seq_pix_limit)
Martin@1
   156
    for bp_per_pix in itertools.count(lower_bound):  # slow linear search
Martin@1
   157
        if tot_seq_pix(seq_sizes, bp_per_pix) <= seq_pix_limit: break
Martin@1
   158
    return bp_per_pix
Martin@1
   159
Martin@1
   160
def get_seq_starts(seq_pix, pix_tween_seqs, margin):
Martin@1
   161
    '''Get the start pixel for each sequence.'''
Martin@1
   162
    seq_starts = []
Martin@1
   163
    pix_tot = margin - pix_tween_seqs
Martin@1
   164
    for i in seq_pix:
Martin@1
   165
        pix_tot += pix_tween_seqs
Martin@1
   166
        seq_starts.append(pix_tot)
Martin@1
   167
        pix_tot += i
Martin@1
   168
    return seq_starts
Martin@1
   169
Martin@645
   170
def get_pix_info(seq_sizes, bp_per_pix, pix_tween_seqs, margin):
Martin@1
   171
    '''Return pixel information about the sequences.'''
Martin@1
   172
    seq_pix = [div_ceil(i, bp_per_pix) for i in seq_sizes]
Martin@1
   173
    seq_starts = get_seq_starts(seq_pix, pix_tween_seqs, margin)
Martin@1
   174
    tot_pix = seq_starts[-1] + seq_pix[-1]
Martin@1
   175
    return seq_pix, seq_starts, tot_pix
Martin@1
   176
Martin@835
   177
def drawLineForward(hits, width, bp_per_pix, beg1, beg2, size):
Martin@639
   178
    while True:
Martin@639
   179
        q1, r1 = divmod(beg1, bp_per_pix)
Martin@639
   180
        q2, r2 = divmod(beg2, bp_per_pix)
Martin@835
   181
        hits[q2 * width + q1] |= 1
Martin@639
   182
        next_pix = min(bp_per_pix - r1, bp_per_pix - r2)
Martin@639
   183
        if next_pix >= size: break
Martin@639
   184
        beg1 += next_pix
Martin@639
   185
        beg2 += next_pix
Martin@639
   186
        size -= next_pix
Martin@639
   187
Martin@835
   188
def drawLineReverse(hits, width, bp_per_pix, beg1, beg2, size):
Martin@639
   189
    beg2 = -1 - beg2
Martin@639
   190
    while True:
Martin@639
   191
        q1, r1 = divmod(beg1, bp_per_pix)
Martin@639
   192
        q2, r2 = divmod(beg2, bp_per_pix)
Martin@835
   193
        hits[q2 * width + q1] |= 2
Martin@639
   194
        next_pix = min(bp_per_pix - r1, r2 + 1)
Martin@639
   195
        if next_pix >= size: break
Martin@639
   196
        beg1 += next_pix
Martin@639
   197
        beg2 -= next_pix
Martin@639
   198
        size -= next_pix
Martin@639
   199
Martin@836
   200
def alignmentPixels(width, height, alignments, bp_per_pix, origins1, origins2):
Martin@640
   201
    hits = [0] * (width * height)  # the image data
Martin@640
   202
    for seq1, seq2, blocks in alignments:
Martin@836
   203
        ori1 = origins1[seq1]
Martin@836
   204
        ori2 = origins2[seq2]
Martin@640
   205
        for beg1, beg2, size in blocks:
Martin@640
   206
            if beg1 < 0:
Martin@640
   207
                beg1 = -(beg1 + size)
Martin@640
   208
                beg2 = -(beg2 + size)
Martin@640
   209
            if beg2 >= 0:
Martin@835
   210
                drawLineForward(hits, width, bp_per_pix,
Martin@835
   211
                                beg1 + ori1, beg2 + ori2, size)
Martin@640
   212
            else:
Martin@835
   213
                drawLineReverse(hits, width, bp_per_pix,
Martin@835
   214
                                beg1 + ori1, beg2 - ori2, size)
Martin@640
   215
    return hits
Martin@1
   216
Martin@650
   217
def expandedSeqDict(seqDict):
Martin@650
   218
    '''Allow lookup by short sequence names, e.g. chr7 as well as hg19.chr7.'''
Martin@650
   219
    newDict = {}
Martin@650
   220
    for name, x in seqDict.items():
Martin@650
   221
        base = name.split(".")[-1]
Martin@650
   222
        newDict[name] = x
Martin@650
   223
        newDict[base] = x
Martin@650
   224
    return newDict
Martin@650
   225
Martin@650
   226
def isExtraFirstGapField(fields):
Martin@650
   227
    return fields[4].isdigit()
Martin@650
   228
Martin@839
   229
def readGaps(fileName, seqLimits):
Martin@650
   230
    '''Read locations of unsequenced gaps, from an agp or gap file.'''
Martin@650
   231
    if not fileName: return
Martin@839
   232
    seqLimits = expandedSeqDict(seqLimits)
Martin@650
   233
    for line in fileinput.input(fileName):
Martin@650
   234
        w = line.split()
Martin@650
   235
        if not w or w[0][0] == "#": continue
Martin@650
   236
        if isExtraFirstGapField(w): w = w[1:]
Martin@650
   237
        if w[4] not in "NU": continue
Martin@650
   238
        seqName = w[0]
Martin@839
   239
        if seqName not in seqLimits: continue
Martin@839
   240
        cropBeg, cropEnd = seqLimits[seqName]
Martin@650
   241
        end = int(w[2])
Martin@650
   242
        beg = end - int(w[5])  # zero-based coordinate
Martin@839
   243
        b = max(beg, cropBeg)
Martin@839
   244
        e = min(end, cropEnd)
Martin@839
   245
        if b >= e: continue
Martin@650
   246
        bridgedText = w[7]
Martin@839
   247
        yield seqName, b, e, bridgedText
Martin@650
   248
Martin@836
   249
def drawUnsequencedGaps(im, gaps, origins, margin, limit, isTop, bridgedText,
Martin@650
   250
                        bp_per_pix, color):
Martin@650
   251
    '''Draw rectangles representing unsequenced gaps.'''
Martin@650
   252
    for seqName, beg, end, b in gaps:
Martin@650
   253
        if b != bridgedText: continue
Martin@836
   254
        ori = origins[seqName]
Martin@836
   255
        b = div_ceil(ori + beg, bp_per_pix)  # use fully-covered pixels only
Martin@836
   256
        e = (ori + end) // bp_per_pix
Martin@650
   257
        if e <= b: continue
Martin@836
   258
        if isTop: box = b, margin, e, limit
Martin@836
   259
        else:     box = margin, b, limit, e
Martin@650
   260
        im.paste(color, box)
Martin@650
   261
Martin@1
   262
def make_label(text, text_size, range_start, range_size):
Martin@1
   263
    '''Return an axis label with endpoint & sort-order information.'''
Martin@1
   264
    text_width  = text_size[0]
Martin@1
   265
    label_start = range_start + (range_size - text_width) // 2
Martin@1
   266
    label_end   = label_start + text_width
Martin@1
   267
    sort_key    = text_width - range_size
Martin@1
   268
    return sort_key, label_start, label_end, text
Martin@1
   269
Martin@645
   270
def get_nonoverlapping_labels(labels, label_space):
Martin@1
   271
    '''Get a subset of non-overlapping axis labels, greedily.'''
Martin@1
   272
    nonoverlapping_labels = []
Martin@1
   273
    for i in labels:
Martin@28
   274
        if True not in [i[1] < j[2] + label_space and j[1] < i[2] + label_space
Martin@28
   275
                        for j in nonoverlapping_labels]:
Martin@1
   276
            nonoverlapping_labels.append(i)
Martin@1
   277
    return nonoverlapping_labels
Martin@1
   278
Martin@837
   279
def get_axis_image(seqNames, name_sizes, seq_starts, seq_pix,
Martin@647
   280
                   font, image_mode, opts):
Martin@1
   281
    '''Make an image of axis labels.'''
Martin@1
   282
    min_pos = seq_starts[0]
Martin@1
   283
    max_pos = seq_starts[-1] + seq_pix[-1]
Martin@28
   284
    height = max(zip(*name_sizes)[1])
Martin@1
   285
    labels = [make_label(i, j, k, l) for i, j, k, l in
Martin@837
   286
              zip(seqNames, name_sizes, seq_starts, seq_pix)]
Martin@1
   287
    labels = [i for i in labels if i[1] >= min_pos and i[2] <= max_pos]
Martin@1
   288
    labels.sort()
Martin@646
   289
    labels = get_nonoverlapping_labels(labels, opts.label_space)
Martin@1
   290
    image_size = max_pos, height
Martin@646
   291
    im = Image.new(image_mode, image_size, opts.border_shade)
Martin@1
   292
    draw = ImageDraw.Draw(im)
Martin@1
   293
    for i in labels:
Martin@1
   294
        position = i[1], 0
Martin@646
   295
        draw.text(position, i[3], font=font, fill=opts.text_color)
Martin@1
   296
    return im
Martin@1
   297
Martin@839
   298
def seqOrigins(seqNames, seq_starts, seqLimits, bp_per_pix):
Martin@837
   299
    for i, j in zip(seqNames, seq_starts):
Martin@839
   300
        yield i, bp_per_pix * j - seqLimits[i][0]
Martin@836
   301
Martin@648
   302
def lastDotplot(opts, args):
Martin@643
   303
    if opts.fontfile:  font = ImageFont.truetype(opts.fontfile, opts.fontsize)
Martin@643
   304
    else:              font = ImageFont.load_default()
Martin@641
   305
Martin@643
   306
    image_mode = 'RGB'
Martin@643
   307
    forward_color = ImageColor.getcolor(opts.forwardcolor, image_mode)
Martin@643
   308
    reverse_color = ImageColor.getcolor(opts.reversecolor, image_mode)
Martin@643
   309
    zipped_colors = zip(forward_color, reverse_color)
Martin@643
   310
    overlap_color = tuple([(i + j) // 2 for i, j in zipped_colors])
Martin@641
   311
Martin@644
   312
    warn("reading alignments...")
Martin@839
   313
    alignments, seqLimits1, seqLimits2 = readAlignments(args[0], opts)
Martin@644
   314
    warn("done")
Martin@641
   315
Martin@649
   316
    if not alignments: raise Exception("there are no alignments")
Martin@641
   317
Martin@839
   318
    seq_info1 = get_seq_info(seqLimits1, font, opts.fontsize, image_mode)
Martin@839
   319
    seq_info2 = get_seq_info(seqLimits2, font, opts.fontsize, image_mode)
Martin@837
   320
    seqNames1, seq_sizes1, name_sizes1, margin1 = seq_info1
Martin@837
   321
    seqNames2, seq_sizes2, name_sizes2, margin2 = seq_info2
Martin@641
   322
Martin@644
   323
    warn("choosing bp per pixel...")
Martin@645
   324
    pix_limit1 = opts.width  - margin1
Martin@645
   325
    pix_limit2 = opts.height - margin2
Martin@646
   326
    bp_per_pix1 = get_bp_per_pix(seq_sizes1, opts.pix_tween_seqs, pix_limit1)
Martin@646
   327
    bp_per_pix2 = get_bp_per_pix(seq_sizes2, opts.pix_tween_seqs, pix_limit2)
Martin@643
   328
    bp_per_pix = max(bp_per_pix1, bp_per_pix2)
Martin@644
   329
    warn("bp per pixel = " + str(bp_per_pix))
Martin@641
   330
Martin@645
   331
    seq_pix1, seq_starts1, width  = get_pix_info(seq_sizes1, bp_per_pix,
Martin@646
   332
                                                 opts.pix_tween_seqs, margin1)
Martin@645
   333
    seq_pix2, seq_starts2, height = get_pix_info(seq_sizes2, bp_per_pix,
Martin@646
   334
                                                 opts.pix_tween_seqs, margin2)
Martin@839
   335
Martin@839
   336
    origins1 = dict(seqOrigins(seqNames1, seq_starts1, seqLimits1, bp_per_pix))
Martin@839
   337
    origins2 = dict(seqOrigins(seqNames2, seq_starts2, seqLimits2, bp_per_pix))
Martin@641
   338
Martin@644
   339
    warn("processing alignments...")
Martin@643
   340
    hits = alignmentPixels(width, height, alignments, bp_per_pix,
Martin@836
   341
                           origins1, origins2)
Martin@644
   342
    warn("done")
Martin@641
   343
Martin@643
   344
    image_size = width, height
Martin@646
   345
    im = Image.new(image_mode, image_size, opts.background_color)
Martin@134
   346
Martin@836
   347
    origins1 = expandedSeqDict(origins1)
Martin@836
   348
    origins2 = expandedSeqDict(origins2)
Martin@839
   349
    gaps1 = list(readGaps(opts.gap1, seqLimits1))
Martin@839
   350
    gaps2 = list(readGaps(opts.gap2, seqLimits2))
Martin@650
   351
    # draw bridged gaps first, then unbridged gaps on top:
Martin@836
   352
    drawUnsequencedGaps(im, gaps1, origins1, margin2, height, True, "yes",
Martin@650
   353
                        bp_per_pix, opts.bridged_color)
Martin@836
   354
    drawUnsequencedGaps(im, gaps2, origins2, margin1, width, False, "yes",
Martin@650
   355
                        bp_per_pix, opts.bridged_color)
Martin@836
   356
    drawUnsequencedGaps(im, gaps1, origins1, margin2, height, True, "no",
Martin@650
   357
                        bp_per_pix, opts.unbridged_color)
Martin@836
   358
    drawUnsequencedGaps(im, gaps2, origins2, margin1, width, False, "no",
Martin@650
   359
                        bp_per_pix, opts.unbridged_color)
Martin@650
   360
Martin@643
   361
    for i in range(height):
Martin@643
   362
        for j in range(width):
Martin@643
   363
            store_value = hits[i * width + j]
Martin@643
   364
            xy = j, i
Martin@643
   365
            if   store_value == 1: im.putpixel(xy, forward_color)
Martin@643
   366
            elif store_value == 2: im.putpixel(xy, reverse_color)
Martin@643
   367
            elif store_value == 3: im.putpixel(xy, overlap_color)
Martin@95
   368
Martin@643
   369
    if opts.fontsize != 0:
Martin@837
   370
        axis1 = get_axis_image(seqNames1, name_sizes1, seq_starts1, seq_pix1,
Martin@647
   371
                               font, image_mode, opts)
Martin@837
   372
        axis2 = get_axis_image(seqNames2, name_sizes2, seq_starts2, seq_pix2,
Martin@647
   373
                               font, image_mode, opts)
Martin@834
   374
        axis2 = axis2.transpose(Image.ROTATE_270)  # !!! bug hotspot
Martin@643
   375
        im.paste(axis1, (0, 0))
Martin@643
   376
        im.paste(axis2, (0, 0))
Martin@1
   377
Martin@643
   378
    for i in seq_starts1[1:]:
Martin@646
   379
        box = i - opts.pix_tween_seqs, margin2, i, height
Martin@646
   380
        im.paste(opts.border_shade, box)
Martin@1
   381
Martin@643
   382
    for i in seq_starts2[1:]:
Martin@646
   383
        box = margin1, i - opts.pix_tween_seqs, width, i
Martin@646
   384
        im.paste(opts.border_shade, box)
Martin@1
   385
Martin@643
   386
    im.save(args[1])
Martin@648
   387
Martin@648
   388
if __name__ == "__main__":
Martin@649
   389
    usage = """%prog --help
Martin@649
   390
   or: %prog [options] maf-or-tab-alignments dotplot.png
Martin@649
   391
   or: %prog [options] maf-or-tab-alignments dotplot.gif
Martin@649
   392
   or: ..."""
Martin@649
   393
    description = "Draw a dotplot of pair-wise sequence alignments in MAF or tabular format."
Martin@649
   394
    op = optparse.OptionParser(usage=usage, description=description)
Martin@651
   395
    op.add_option("-1", "--seq1", metavar="PATTERN", action="append",
Martin@651
   396
                  help="which sequences to show from the 1st genome")
Martin@651
   397
    op.add_option("-2", "--seq2", metavar="PATTERN", action="append",
Martin@651
   398
                  help="which sequences to show from the 2nd genome")
Martin@648
   399
    # Replace "width" & "height" with a single "length" option?
Martin@648
   400
    op.add_option("-x", "--width", type="int", default=1000,
Martin@648
   401
                  help="maximum width in pixels (default: %default)")
Martin@648
   402
    op.add_option("-y", "--height", type="int", default=1000,
Martin@648
   403
                  help="maximum height in pixels (default: %default)")
Martin@649
   404
    op.add_option("-f", "--fontfile", metavar="FILE",
Martin@648
   405
                  help="TrueType or OpenType font file")
Martin@649
   406
    op.add_option("-s", "--fontsize", metavar="SIZE", type="int", default=11,
Martin@648
   407
                  help="TrueType or OpenType font size (default: %default)")
Martin@649
   408
    op.add_option("-c", "--forwardcolor", metavar="COLOR", default="red",
Martin@649
   409
                  help="color for forward alignments (default: %default)")
Martin@649
   410
    op.add_option("-r", "--reversecolor", metavar="COLOR", default="blue",
Martin@649
   411
                  help="color for reverse alignments (default: %default)")
Martin@839
   412
    op.add_option("--trim1", action="store_true",
Martin@839
   413
                  help="trim unaligned sequence flanks from the 1st genome")
Martin@839
   414
    op.add_option("--trim2", action="store_true",
Martin@839
   415
                  help="trim unaligned sequence flanks from the 2nd genome")
Martin@650
   416
    og = optparse.OptionGroup(op, "Unsequenced gap options")
Martin@650
   417
    og.add_option("--gap1", metavar="FILE",
Martin@650
   418
                  help="read genome1 unsequenced gaps from agp or gap file")
Martin@650
   419
    og.add_option("--gap2", metavar="FILE",
Martin@650
   420
                  help="read genome2 unsequenced gaps from agp or gap file")
Martin@650
   421
    og.add_option("--bridged-color", metavar="COLOR", default="yellow",
Martin@650
   422
                  help="color for bridged gaps (default: %default)")
Martin@650
   423
    og.add_option("--unbridged-color", metavar="COLOR", default="pink",
Martin@650
   424
                  help="color for unbridged gaps (default: %default)")
Martin@650
   425
    op.add_option_group(og)
Martin@648
   426
    (opts, args) = op.parse_args()
Martin@648
   427
    if len(args) != 2: op.error("2 arguments needed")
Martin@648
   428
Martin@648
   429
    opts.text_color = "black"
Martin@648
   430
    opts.background_color = "white"
Martin@648
   431
    opts.pix_tween_seqs = 2  # number of border pixels between sequences
Martin@648
   432
    opts.border_shade = 239, 239, 239  # the shade of grey for border pixels
Martin@648
   433
    opts.label_space = 5     # minimum number of pixels between axis labels
Martin@648
   434
Martin@649
   435
    try: lastDotplot(opts, args)
Martin@649
   436
    except KeyboardInterrupt: pass  # avoid silly error message
Martin@649
   437
    except Exception, e:
Martin@649
   438
        prog = os.path.basename(sys.argv[0])
Martin@649
   439
        sys.exit(prog + ": error: " + str(e))