def parse_pfam(infile, color=None, y=-2.5, size=8):
entities = []
spans = []
with open(infile) as f:
for l in f:
if l.startswith('#'): continue
elif not l.strip(): continue
else:
sl = l.strip().split()
#label = l[181:].strip()
#start = int(l[152:157].strip())
#end = int(l[158:163].strip())
#label = sl[-1]
label = sl[1]
start = int(sl[19])
end = int(sl[20])
dy = 0
for span in spans:
if (span[0] <= start <= span[1]) or (span[0] <= end <=
span[1]):
dy = 0.3
entities.append(
quod.Region([[start, end]], [y - 0.15 + dy, 0.15],
label,
style=color,
size=size))
spans.append([start, end])
return entities
def parse_pfam(infile, styledict, color=None, y=-2.75, size=8):
entities = []
spans = []
with open(infile) as f:
for l in f:
if l.startswith('#'): continue
elif not l.strip(): continue
else:
sl = l.strip().split()
#label = l[181:].strip()
#start = int(l[152:157].strip())
#end = int(l[158:163].strip())
#label = sl[-1]
domain = sl[1]
#TODO: pass clandict around to speed up clan searches
clan = get_clan(domain)
style = 3
if clan.strip():
if clan in styledict: style = styledict[clan]
else:
if styledict: style = max(styledict.values()) + 1
style = 3
styledict[clan] = style
elif domain in styledict:
style = styledict[domain]
else:
if styledict: style = max(styledict.values()) + 1
else: style = 3
styledict[domain] = style
#print(styledict)
start = int(sl[19])
end = int(sl[20])
dy = 0
ystep = 0.7
for span in spans:
if (span[0] <= start <= span[1]) or (span[0] <= end <=
span[1]):
dy += ystep
entities.append(
quod.Region([[start, end]], [y - 0.2 + dy, 0.2],
domain,
style=style,
size=size))
#entities.append(quod.Region([[start, end]], [y-0.2+dy+ystep, 0.2], domain, style=style, size=size))
#entities.append(quod.Region([[start, end]], [y-0.2+dy+ystep*2, 0.2], domain, style=style, size=size))
#entities.append(quod.Region([[start, end]], [y-0.2+dy+ystep*3, 0.2], domain, style=style, size=size))
#entities.append(quod.Region([[start, end]], [y-0.2+dy+ystep*4, 0.2], domain, style=style, size=size))
spans.append([start, end])
return entities, styledict
def run_quod(cfg):
entities = {}
maxlens = {}
for name in cfg['subplots']:
entities[name] = []
maxlens[name] = 0
for seq in cfg['sequences'][name]:
entities[name].append(quod.What(seq))
maxlens[name] = max(maxlens[name], len(entities[name][-1]))
#walls
if name in cfg['walls']:
for walldef in cfg['walls'][name]:
entities[name].append(quod.Wall(spans=[[walldef['start'], walldef['end']]], y=walldef['y'], ylim=walldef['lim'], thickness=walldef['thickness'], wedge=walldef['thickness']))
#domains
if name in cfg['domains']:
for domaindef in cfg['domains'][name]:
entities[name].append(quod.Region(
spans=domaindef['xlim'],
yspan=domaindef['ylim'],
label=domaindef['label'],
style=domaindef['color'],
pos=domaindef['pos'],
size=domaindef['fontsize'],
center=domaindef['halign']
))
cfg['weights'][name] *= maxlens[name]
if 'tms' in cfg: do_tms_stuff(entities, cfg)
fig = Figure()
canvas = FigureCanvas(fig)
fig.set_tight_layout(True)
gs = gridspec.GridSpec(len(cfg['rowscheme']), cfg['hres'])
hgap = cfg['hgap']/cfg['width']/2
margin = 0.03 if 'margin' not in cfg else cfg['margin']
gs.update(
left=margin + hgap,
right=cfg['width'] - margin,
top=cfg['height'] - margin-hgap,
bottom=margin, wspace=0)
def get_limit(wt1, wtlist, offset=0):
x = int((offset + wt1/sum(wtlist)) * cfg['hres'])
return max(0, min(x, cfg['hres']))
if cfg['mode'] == 'hvordan':
lims = []
cfg['lims'] = {}
for namelist in cfg['rowscheme']:
row = [cfg['weights'][name] for name in namelist]
last = -hgap
if len(namelist) == 1:
cfg['lims'][namelist[0]] = [0, cfg['hres']]
else:
bounds = [0]
s = 0
rawweights = [cfg['weights'][name] for name in namelist]
weights = [w/sum(rawweights) for w in rawweights]
for i in range(len(namelist) - 1):
bounds.append((weights[i] - hgap) * cfg['hres'])
bounds.append((weights[i] + hgap) * cfg['hres'])
#for i in namelist
bounds.append(cfg['hres'])
bounds = [int(x) for x in bounds]
for i in range(0, len(namelist)):
name = namelist[i]
cfg['lims'][name] = [bounds[2*i], bounds[2*i + 1]]
axdict = {}
for r, row in enumerate(cfg['rowscheme']):
for name in row:
#axdict[name] = fig.add_subplot(gs[r, cfg['lims'][name][0]:cfg['lims'][name][1]])
axdict[name] = fig.add_subplot(gs[r,cfg['lims'][name][0]:cfg['lims'][name][1]])
n = 0
name2row = {}
stretchlabel = {}
firstcol = set()
for r, row in enumerate(cfg['rowscheme']):
for c, name in enumerate(row):
if not c: firstcol.add(name)
name2row[name] = r
if 'ylabel' in cfg and name in cfg['ylabel']: ylabel = cfg['ylabel'][name]
else: ylabel = '' if c else None
yticks = [] if c else None
plot(cfg, name, fig, canvas, axdict[name], entities[name], ylabel=ylabel, yticks=yticks)
try: title = cfg['title'][name]
except KeyError: title = chr(65 + n)
axdict[name].set_title(title, loc='left')
n += 1
if len(row) > 1:
ax = fig.add_subplot(gs[r,:])
ax.set_xticks([0])
ax.axes.get_yaxis().set_visible(False)
ax.set_frame_on(False)
#TODO: expose customizing xlabels per-plot/per-row
ax.set_xlabel('Position')
ax.tick_params(labelcolor=(0,0,0,0), color=(0,0,0,0))
stretchlabel[r] = ax
elif len(row) == 1:
#ax = fig.add_subplot(gs[r,:])
ax = axdict[row[0]]
ax.set_xlabel('Position')
stretchlabel[r] = ax
if 'xlabel' in cfg:
for name in cfg['xlabel']:
stretchlabel[name2row[name]].set_xlabel(cfg['xlabel'][name])
if 'ylabel' in cfg:
for name in cfg['ylabel']:
stretchlabel[name2row[name]].set_ylabel(cfg['ylabel'][name])
if 'font' in cfg:
for name in cfg['font']:
for target in cfg['font'][name]:
if target.endswith('ticks'):
axdict[name].tick_params(labelsize=cfg['font'][name][target])
if name in firstcol:
stretchlabel[name2row[name]].tick_params(labelsize=cfg['font'][name][target])
elif target.endswith('xaxis'):
axdict[name].set_xlabel(axdict[name].get_xlabel(),
fontsize=cfg['font'][name][target])
stretchlabel[name2row[name]].set_xlabel(stretchlabel[name2row[name]].get_xlabel(),
fontsize=cfg['font'][name][target])
elif target.endswith('yaxis'):
axdict[name].set_ylabel(axdict[name].get_ylabel(),
fontsize=cfg['font'][name][target])
elif target == 'title':
axdict[name].set_title(axdict[name].get_title(),
fontsize=cfg['font'][name][target])
#gs.tight_layout(fig, pad=cfg['margin'], w_pad=cfg['hgap'], h_pad=0.0)
gs.tight_layout(fig, pad=0, w_pad=cfg['hgap'], h_pad=cfg['vgap'])
fig.savefig(cfg['outfile'], dpi=cfg['dpi'])
else: raise NotImplementedError('Unimplemented mode: {}'.format(cfg['mode']))
def get_domain(self): '''return an object consisting of some text and a strongly colored box''' y = -2 if self.y is None else self.y return quod.Region(spans=[[self.left, self.right]], yspan=[y-0.15, 0.15], label=self.label, style=self.style)
def run_quod(cfg):
entities = {}
maxlens = {}
for name in cfg['subplots']:
entities[name] = []
maxlens[name] = 0
mode = cfg['mode'][name]
if mode in ('hvordan', 'quod'):
if cfg['frag'][name]:
for i in range(0, len(cfg['sequences'][name]), 2):
entities[name].append(
quod.FragmentWhat(cfg['sequences'][name][i],
cfg['sequences'][name][i + 1]))
maxlens[name] = max(maxlens[name], len(entities[name][-1]))
else:
for seq in cfg['sequences'][name]:
entities[name].append(quod.What(seq))
maxlens[name] = max(maxlens[name], len(entities[name][-1]))
elif mode in ('avehas', ):
if avehas3 is None: raise ImportError('Could not find avehas3.py')
for seq in cfg['sequences'][name]:
entities[name].append(
avehas3.Avehas(io.BytesIO(seq.encode('utf-8')),
amphi=cfg['amphi'][name]))
#print(name, seq, maxlens[name])
#for msa in avehas3.AlignIO.parse(
#print(len(entities[name][-1]))
maxlens[name] = max(maxlens[name], len(entities[name][-1]))
elif mode in ('empty', ):
spacer = Spacer(length=cfg['baselength'][name])
maxlens[name] = cfg['baselength'][name]
entities[name].append(spacer)
for seq in cfg['sequences'][name]:
#walls
if name in cfg['walls']:
for walldef in cfg['walls'][name]:
entities[name].append(
quod.Wall(spans=[[walldef['start'], walldef['end']]],
y=walldef['y'],
ylim=walldef['lim'],
thickness=walldef['thickness'],
wedge=walldef['thickness']))
#domains
if name in cfg['domains']:
for domaindef in cfg['domains'][name]:
entities[name].append(
quod.Region(spans=domaindef['xlim'],
yspan=domaindef['ylim'],
label=domaindef['label'],
style=domaindef['color'],
pos=domaindef['pos'],
size=domaindef['fontsize'],
center=domaindef['halign']))
if name in cfg['text']:
label = cfg['text'][name]
entities[name].append(
quod.Text(label['pos'], label['text'],
label.get('halign', 'l'), label.get('font', None)))
if cfg['baselength'][name] is None:
cfg['baselength'][name] = maxlens[name]
#cfg['length'][name] = maxlens[name]
#cfg['weight'][name] = maxlens[name]
cfg['length'][name] = cfg['baselength'][name] * cfg['weight'][name]
if 'tms' in cfg: do_tms_stuff(entities, cfg)
fig = quod.Figure()
fig.set_tight_layout(True)
gs = gridspec.GridSpec(len(cfg['rowscheme']), cfg['hres'])
hgap = cfg['hgap'] / cfg['width'] / 2
margin = 0.03 if 'margin' not in cfg else cfg['margin']
gs.update(left=margin + hgap,
right=cfg['width'] - margin,
top=cfg['height'] - margin - hgap,
bottom=margin,
wspace=0)
def get_limit(wt1, wtlist, offset=0):
x = int((offset + wt1 / sum(wtlist)) * cfg['hres'])
return max(0, min(x, cfg['hres']))
if True:
lims = []
cfg['lims'] = {}
for namelist in cfg['rowscheme']:
row = [cfg['length'][name] for name in namelist]
last = -hgap
if len(namelist) == 1:
cfg['lims'][namelist[0]] = [0, cfg['hres']]
else:
bounds = [0]
s = 0
rawweight = [cfg['length'][name] for name in namelist]
weight = [w / sum(rawweight) for w in rawweight]
for i in range(len(namelist) - 1):
bounds.append((weight[i] - hgap) * cfg['hres'])
bounds.append((weight[i] + hgap) * cfg['hres'])
#for i in namelist
bounds.append(cfg['hres'])
bounds = [int(x) for x in bounds]
for i in range(0, len(namelist)):
name = namelist[i]
cfg['lims'][name] = [bounds[2 * i], bounds[2 * i + 1]]
axdict = {}
for r, row in enumerate(cfg['rowscheme']):
for name in row:
#axdict[name] = fig.add_subplot(gs[r, cfg['lims'][name][0]:cfg['lims'][name][1]])
axdict[name] = fig.add_subplot(
gs[r, cfg['lims'][name][0]:cfg['lims'][name][1]])
n = 0
name2row = {}
stretchlabel = {}
firstcol = set()
for r, row in enumerate(cfg['rowscheme']):
for c, name in enumerate(row):
if not c: firstcol.add(name)
name2row[name] = r
if 'ylabel' in cfg and name in cfg['ylabel']:
ylabel = cfg['ylabel'][name]
else:
ylabel = '' if c else None
yticks = None #[] if c else None
plot(cfg,
name,
fig,
axdict[name],
entities[name],
ylabel=ylabel,
yticks=yticks)
if c:
axdict[name].set_yticklabels(
[] * len(axdict[name].get_yticklabels()))
try:
title = cfg['title'][name]
except KeyError:
title = chr(65 + n)
titlefont = get_titlefont(cfg, name)
#TODO: implement rowtitle separately
#axdict[name].set_title(cfg['title'].get(name, ''), loc='left', fontsize=titlefont)
titlepad = quod.matplotlib.rcParams['axes.titlepad']
transOffset = matplotlib.transforms.ScaledTranslation(
0., titlepad / 72., fig.dpi_scale_trans)
if name in cfg['ctitle']:
subtitlefont = get_titlefont(cfg,
name,
subname='ctitle',
default=titlefont)
t = axdict[name].text(0.5,
1.0,
cfg['ctitle'].get(name, ''),
ha='center',
va='baseline',
fontsize=subtitlefont,
transform=axdict[name].transAxes)
t.set_transform(axdict[name].transAxes + transOffset)
if name in cfg['ltitle']:
subtitlefont = get_titlefont(cfg,
name,
subname='ltitle',
default=titlefont)
t = axdict[name].text(0.0,
1.0,
cfg['ltitle'].get(name, ''),
ha='left',
va='baseline',
fontsize=subtitlefont,
transform=axdict[name].transAxes)
t.set_transform(axdict[name].transAxes + transOffset)
if name in cfg['rtitle']:
subtitlefont = get_titlefont(cfg,
name,
subname='rtitle',
default=titlefont)
t = axdict[name].text(1.0,
1.0,
cfg['rtitle'].get(name, ''),
ha='right',
va='baseline',
fontsize=subtitlefont,
transform=axdict[name].transAxes)
t.set_transform(axdict[name].transAxes + transOffset)
if name in cfg['title']:
axdict[name].set_title(cfg['title'].get(name, ''),
loc='left',
fontsize=titlefont)
n += 1
if len(row) > 1:
ax = fig.add_subplot(gs[r, :])
ax.set_xticks([0])
ax.axes.get_yaxis().set_visible(False)
#ax.axes.get_xaxis().set_visible(False)
ax.set_frame_on(False)
#TODO: expose customizing xlabels per-plot/per-row
ax.set_xlabel('Position')
ax.tick_params(labelcolor=(1, 1, 1, 0), color=(1, 1, 1, 0))
for side in ['left', 'right', 'bottom', 'top']:
ax.spines[side].set_color((1, 1, 1, 0))
ax.spines[side].set_linewidth(0)
stretchlabel[r] = ax
elif len(row) == 1:
#ax = fig.add_subplot(gs[r,:])
ax = axdict[row[0]]
ax.set_xlabel('Position')
stretchlabel[r] = ax
if 'xlabel' in cfg:
for name in cfg['xlabel']:
if name.startswith('row:'):
stretchlabel[name2row[name[4:]]].set_xlabel(
cfg['xlabel'][name])
else:
axdict[name].set_xlabel(cfg['xlabel'][name])
stretchlabel[name2row[name]].set_xlabel(' ')
if 'ylabel' in cfg:
for name in cfg['ylabel']:
stretchlabel[name2row[name]].set_ylabel(cfg['ylabel'][name])
if 'font' in cfg:
for name in cfg['font']:
for target in cfg['font'][name]:
if target.endswith('ticks'):
axdict[name].tick_params(
labelsize=cfg['font'][name][target])
if name in firstcol:
stretchlabel[name2row[name]].tick_params(
labelsize=cfg['font'][name][target])
elif target.endswith('xaxis'):
axdict[name].set_xlabel(
axdict[name].get_xlabel(),
fontsize=cfg['font'][name][target])
stretchlabel[name2row[name]].set_xlabel(
stretchlabel[name2row[name]].get_xlabel(),
fontsize=cfg['font'][name][target])
elif target.endswith('yaxis'):
axdict[name].set_ylabel(
axdict[name].get_ylabel(),
fontsize=cfg['font'][name][target])
#gs.tight_layout(fig, pad=cfg['margin'], w_pad=cfg['hgap'], h_pad=0.0)
#gs.tight_layout(fig, pad=0, w_pad=cfg['hgap'], h_pad=cfg['vgap'])
gs.tight_layout(fig,
pad=cfg['margin'],
w_pad=cfg['hgap'],
h_pad=cfg['vgap'])
fig.savefig(cfg['outfile'], dpi=cfg['dpi'])
else:
raise NotImplementedError('Unimplemented mode: {}'.format(cfg['mode']))