def layout_raw(node: TreeNode, tight_mode: bool = True) -> None:
"""Layout implementation for a tree node
Parameters
----------
node : TreeNode
the root of the taxonomy tree / sub-tree
tight_mode : bool, default=True
a mode to print node names more tightly
Returns
-------
None
"""
if tight_mode:
name_segments = node.name.split(' ')
for i, name_segment in enumerate(name_segments):
name_face = TextFace(name_segment, tight_text=True)
name_face.rotation = 270
node.add_face(name_face, column=i, position="branch-right")
else:
name_face = TextFace(node.name, tight_text=True)
name_face.rotation = 270
node.add_face(name_face, column=0, position="branch-right")
nst = NodeStyle()
nst["fgcolor"] = "black"
nst["size"] = 20
nst["shape"] = "circle"
node.set_style(nst)
def my_layout(node): #coloca o nome em cada nó
node.name
F = TextFace(
node.name.replace("*", "\n"),
tight_text=True) #substitui onde tem estrela p quebra de linha
add_face_to_node(F, node, column=0, position="branch-right")
F.rotation = -90 #rotação do nome no nó
def layout(node):
# Make node name empty for particular nodes
# if node.name in ('information retrieval', 'information systems'):
# node.name = ''
# x = 0
#
# if node.name == '3 items':
# node.name = ''
# Some long name war here
try:
print_label = int(
node.e
) < 3 or node.Hd == '1' or node.Of == '1' or node.Gap == '1' or node.ForceLabel == '1'
if print_label:
name_split = node.name.split('|')
column = 0
for line in name_split:
tw = textwrap.TextWrapper(width=20)
names = tw.wrap(line)
for n in names:
short_name = TextFace(n, tight_text=True)
short_name.rotation = 270
node.add_face(short_name,
column=column,
position="branch-right")
column += 1
# Create and apply node style
nst = NodeStyle()
if .4 >= float(node.u) > 0:
nst["fgcolor"] = "#90ee90"
elif .6 >= float(node.u) > .4:
nst["fgcolor"] = "green"
elif float(node.u) > .6:
nst["fgcolor"] = "#004000"
elif node.Gap == '1':
nst["fgcolor"] = "red"
else:
nst["fgcolor"] = NO_SUPPORT_COLOR
if node.Hd == '1' or node.Of == '1':
nst["size"] = 40
nst["shape"] = 'square'
else:
nst["size"] = 20
nst["shape"] = 'circle'
# if node.Sq == '1' and float(node.u) > 0:
# nst["shape"] = 'square'
node.set_style(nst)
except:
print(f'Exception at {node}')
def add_text_face(self,
taxon2text,
header_name,
color_scale=False):
from metagenlab_libs.colors import get_categorical_color_scale
if color_scale:
value2color = get_categorical_color_scale(taxon2text.values())
self._add_header(header_name)
# add column
for i, lf in enumerate(self.tree.iter_leaves()):
if lf.name in taxon2text:
n = TextFace('%s' % taxon2text[lf.name])
if color_scale:
n.background.color = value2color[taxon2text[lf.name]]
else:
print(lf.name, "not in", taxon2text)
n = TextFace('-')
n.margin_top = 1
n.margin_right = 10
n.margin_left = 10
n.margin_bottom = 1
n.opacity = 1.
if self.rotate:
n.rotation= 270
lf.add_face(n, self.column_count, position="aligned")
self.column_count += 1
def attribute_legend(node, authors, styles):
indx1 = authors.index(node.name)
author_styles = styles[indx1]
text = " "
counter = 0
for style in author_styles:
if counter % 2 == 0:
text += " \n "
counter += 1
text += style + ", "
text = text[:-2]
text = " " + text
#F = TextFace(node.name, tight_text=True, fsize=15, fgcolor="white")
#add_face_to_node(F, node, column=0, position="branch-right")
N = TextFace(text,
fgcolor="Black",
fsize=21,
fstyle="bold",
bold=False,
tight_text=False)
if not node.is_leaf():
N.rotation = 90
Nspace = TextFace(" ", fgcolor="Black", fsize=21, bold=True)
node.add_face(face=Nspace, column=1)
node.add_face(face=N, column=2)
def my_layout(node):
N = None
# If user specified labels are specified add these:
if 'node_label' in tree_features:
for t in tree_features['node_label']:
node_attr = ifhasthenget(node, t)
if node_attr and to_string(node_attr):
if t == 'name' and hasattr(node, 'cl'):
continue
textface = to_string(node_attr)
N = TextFace(textface, fsize=12, fgcolor='black')
if 'e-' in textface and to_float(
textface
): # Stupid exception because ete3 rendering doesn't understand scientific notation
N = TextFace('%2.1e ' % to_float(textface),
fsize=12,
fgcolor='black')
# Add default values:
elif not no_default and node.frequency > 1 and not hasattr(node, 'cl'):
N = TextFace(
node.frequency, fsize=12,
fgcolor='black') # Default: use node frequency as TextFace
if N is not None:
N.rotation = -90
faces.add_face_to_node(N, node, 0, position='branch-top')
def styleFace(val):
x = TextFace(val)
x.margin_bottom = 5
x.margin_right = 10
x.rotation = 270
x.fsize = 6
return x
def to_tree_node(self):
t = Tree(f"{self.function_type};", format=1)
for child in self.children:
t.add_child(child.to_tree_node())
tf = TextFace(f"{self.function_type}")
tf.rotation = -90
t.add_face(tf, column=1, position="branch-top")
return t
def rotation_layout(node):
if node.is_leaf():
if node.name == 'X':
F = TextFace(node.name, tight_text=True, fgcolor='Blue')
F.rotation = 90
add_face_to_node(F, node, column=0, position="branch-right")
elif node.name == 'Y':
F = TextFace(node.name, tight_text=True, fgcolor='Red')
F.rotation = 90
add_face_to_node(F, node, column=0, position="branch-right")
elif node.name == 'A':
F = TextFace("")
add_face_to_node(F, node, column=0, position="branch-right")
else:
F = TextFace(node.name, tight_text=True, fgcolor='Green')
F.rotation = 90
add_face_to_node(F, node, column=0, position="branch-right")
def my_layout(node):
F = TextFace("\t" + node.name,
tight_text=True,
fstyle="bold",
fsize=40,
fgcolor="black")
F.rotation = 90
add_face_to_node(F, node, column=0, position="branch-bottom")
def my_layout(node):
F = TextFace(node.name, tight_text=True)
F.fsize=6
F.margin_left=5
F.margin_right=5
F.margin_top=0
F.margin_bottom=15
F.rotation=-90
add_face_to_node(F, node, column=0, position="branch-right")
def to_tree_node(self):
t = Tree("IF;", format=1)
t.add_child(self.children[1].to_tree_node())
t.add_child(self.children[0].to_tree_node())
t.add_child(self.children[2].to_tree_node())
tf = TextFace("IF")
tf.rotation = -90
t.add_face(tf, column=1, position="branch-top")
return t
def rotation_layout(node):
if node.is_leaf():
F = TextFace(node.name, tight_text=True)
F.rotation = randint(0, 360)
add_face_to_node(TextFace("third" ), node, column=8, position="branch-right")
add_face_to_node(TextFace("second" ), node, column=2, position="branch-right")
add_face_to_node(F, node, column=0, position="branch-right")
F.border.width = 1
F.inner_border.width = 1
def my_layout(node):
circle_color = 'lightgray' if colormap is None or node.name not in colormap else colormap[
node.name]
text_color = 'black'
if isinstance(circle_color, str):
C = CircleFace(radius=max(3, 10 * scipy.sqrt(node.frequency)),
color=circle_color,
label={
'text': str(node.frequency),
'color': text_color
} if node.frequency > 0 else None)
C.rotation = -90
C.hz_align = 1
faces.add_face_to_node(C, node, 0)
else:
P = PieChartFace(
[100 * x / node.frequency for x in circle_color.values()],
2 * 10 * scipy.sqrt(node.frequency),
2 * 10 * scipy.sqrt(node.frequency),
colors=[(color if color != 'None' else 'lightgray')
for color in list(circle_color.keys())],
line_color=None)
T = TextFace(' '.join(
[str(x) for x in list(circle_color.values())]),
tight_text=True)
T.hz_align = 1
T.rotation = -90
faces.add_face_to_node(P, node, 0, position='branch-right')
faces.add_face_to_node(T, node, 1, position='branch-right')
if idlabel:
T = TextFace(node.name, tight_text=True, fsize=6)
T.rotation = -90
T.hz_align = 1
faces.add_face_to_node(
T,
node,
1 if isinstance(circle_color, str) else 2,
position='branch-right')
def _add_header(self,
header_name,
column_add=0):
n = TextFace(f'{header_name}')
n.margin_top = 1
n.margin_right = 1
n.margin_left = 20
n.margin_bottom = 1
n.hz_align = 2
n.vt_align = 2
n.rotation = 270
n.inner_background.color = "white"
n.opacity = 1.
# add header
self.tss.aligned_header.add_face(n, self.column_count-1+column_add)
def layout_lift(node: TreeNode, levels: int = 3) -> None:
"""Layout implementation for a tree node
Parameters
----------
node : TreeNode
the root of the taxonomy tree / sub-tree
levels : int
a number of tree levels to draw
Returns
-------
None
"""
name = TextFace(node.name if
(int(node.e) < levels or node.Hd == "1") else "",
tight_text=True)
name.rotation = 270
node.add_face(name, column=0, position="branch-right")
nst = NodeStyle()
if .2 >= float(node.u) > 0:
nst["fgcolor"] = "#90ee90"
elif .4 >= float(node.u) > .2:
nst["fgcolor"] = "green"
elif float(node.u) > .4:
nst["fgcolor"] = "#004000"
else:
nst["fgcolor"] = "red"
if node.Hd == "0":
nst["size"] = 20
nst["shape"] = "square"
else:
if node.Ch == "1":
nst["size"] = 40
nst["shape"] = "circle"
else:
nst["size"] = 40
nst["shape"] = "circle"
if node.Sq == "1":
nst["shape"] = "circle"
node.set_style(nst)
def add_heatmap(self,
taxon2value,
header_name,
continuous_scale=False,
show_text=False):
from metagenlab_libs.colors import get_continuous_scale
self._add_header(header_name)
if continuous_scale:
color_scale = get_continuous_scale(taxon2value.values())
for i, lf in enumerate(self.tree.iter_leaves()):
if not lf.name in taxon2value:
n = TextFace('')
else:
value = taxon2value[lf.name]
if show_text:
n = TextFace('%s' % value)
else:
n = TextFace(' ')
n.margin_top = 2
n.margin_right = 3
n.margin_left = 3
n.margin_bottom = 2
n.hz_align = 1
n.vt_align = 1
n.border.width = 3
n.border.color = "#ffffff"
if continuous_scale:
n.background.color = rgb2hex(color_scale[0].to_rgba(float(value)))
n.opacity = 1.
i+=1
if self.rotate:
n.rotation = 270
lf.add_face(n, self.column_count, position="aligned")
self.column_count += 1
def add_heatmap(self,
taxon2value,
header_name,
scale_type="continuous",
palette=False):
from metagenlab_libs.colors import get_categorical_color_scale
from metagenlab_libs.colors import get_continuous_scale
if scale_type == "continuous":
scale = get_continuous_scale(taxon2value.values())
self.add_continuous_colorscale_legend("Closest hit identity",
min(taxon2value.values()),
max(taxon2value.values()),
scale)
elif scale_type == "categorical":
scale = get_categorical_color_scale(taxon2value.values())
self.add_categorical_colorscale_legend("MLST",
scale)
else:
raise IOError("unknown type")
for i, lf in enumerate(self.tree.iter_leaves()):
n = TextFace(" " * int(self.text_scale))
if lf.name in taxon2value:
value = taxon2value[lf.name]
n = TextFace(" " * int(self.text_scale))
if scale_type == "categorical":
n.inner_background.color = scale[value]
if scale_type == "continuous":
n.inner_background.color = rgb2hex(scale[0].to_rgba(float(value)))
n.margin_top = 0
n.margin_right = 0
n.margin_left = 10
n.margin_bottom = 0
n.opacity = 1.
if self.rotate:
n.rotation= 270
lf.add_face(n, self.column_count, position="aligned")
self.column_count += 1
def add_simple_barplot(self,
taxon2value,
header_name,
color=False,
show_values=False,
substract_min=False,
max_value=False):
print("scale factor", self.text_scale)
if not show_values:
self._add_header(header_name, column_add=0)
else:
self._add_header(header_name, column_add=1)
values_lists = [float(i) for i in taxon2value.values()]
min_value = min(values_lists)
if substract_min:
values_lists = [i-min_value for i in values_lists]
for taxon in list(taxon2value.keys()):
taxon2value[taxon] = taxon2value[taxon]-min_value
if not color:
color = self._get_default_barplot_color()
for i, lf in enumerate(self.tree.iter_leaves()):
try:
value = taxon2value[lf.name]
except:
value = 0
if show_values:
barplot_column = 1
if isinstance(value, float):
a = TextFace(" %s " % str(round(value,2)))
else:
a = TextFace(" %s " % str(value))
a.margin_top = 1
a.margin_right = 2
a.margin_left = 5
a.margin_bottom = 1
if self.rotate:
a.rotation = 270
lf.add_face(a, self.column_count, position="aligned")
else:
barplot_column = 0
if not max_value:
fraction_biggest = (float(value)/max(values_lists))*100
else:
fraction_biggest = (float(value)/max_value)*100
fraction_rest = 100-fraction_biggest
b = StackedBarFace([fraction_biggest, fraction_rest],
width=100 * (self.text_scale/3),
height=18,
colors=[color, 'white'])
b.rotation= 0
#b.inner_border.color = "grey"
#b.inner_border.width = 0
b.margin_right = 10
b.margin_left = 10
b.hz_align = 2
b.vt_align = 2
b.rotable = False
if self.rotate:
b.rotation = 270
lf.add_face(b, self.column_count + barplot_column, position="aligned")
self.column_count += (1 + barplot_column)
def plot_tree_stacked_barplot(
tree_file,
taxon2value_list_barplot=False,
header_list=False, # header stackedbarplots
taxon2set2value_heatmap=False,
taxon2label=False,
header_list2=False, # header counts columns
biodb=False,
column_scale=True,
general_max=False,
header_list3=False,
set2taxon2value_list_simple_barplot=False,
set2taxon2value_list_simple_barplot_counts=True,
rotate=False,
taxon2description=False):
'''
taxon2value_list_barplot list of lists:
[[bar1_part1, bar1_part2,...],[bar2_part1, bar2_part2]]
valeures de chaque liste transformes en pourcentages
:param tree_file:
:param taxon2value_list:
:param biodb:
:param exclude_outgroup:
:param bw_scale:
:return:
'''
if biodb:
from chlamdb.biosqldb import manipulate_biosqldb
server, db = manipulate_biosqldb.load_db(biodb)
taxon2description = manipulate_biosqldb.taxon_id2genome_description(
server, biodb, filter_names=True)
t1 = Tree(tree_file)
# Calculate the midpoint node
R = t1.get_midpoint_outgroup()
# and set it as tree outgroup
t1.set_outgroup(R)
colors2 = [
"red", "#FFFF00", "#58FA58", "#819FF7", "#F781F3", "#2E2E2E",
"#F7F8E0", 'black'
]
colors = [
"#7fc97f", "#386cb0", "#fdc086", "#ffffb3", "#fdb462", "#f0027f",
"#F7F8E0", 'black'
] # fdc086ff 386cb0ff f0027fff
tss = TreeStyle()
tss.draw_guiding_lines = True
tss.guiding_lines_color = "gray"
tss.show_leaf_name = False
if column_scale and header_list2:
import matplotlib.cm as cm
from matplotlib.colors import rgb2hex
import matplotlib as mpl
column2scale = {}
col_n = 0
for column in header_list2:
values = taxon2set2value_heatmap[column].values()
#print values
if min(values) == max(values):
min_val = 0
max_val = 1.5 * max(values)
else:
min_val = min(values)
max_val = max(values)
#print 'min-max', min_val, max_val
norm = mpl.colors.Normalize(vmin=min_val, vmax=max_val) # *1.1
if col_n < 4:
cmap = cm.OrRd #
else:
cmap = cm.YlGnBu #PuBu#OrRd
m = cm.ScalarMappable(norm=norm, cmap=cmap)
column2scale[column] = [m, float(max_val)] # *0.7
col_n += 1
for i, lf in enumerate(t1.iter_leaves()):
#if taxon2description[lf.name] == 'Pirellula staleyi DSM 6068':
# lf.name = 'Pirellula staleyi DSM 6068'
# continue
if i == 0:
if taxon2label:
n = TextFace(' ')
n.margin_top = 1
n.margin_right = 1
n.margin_left = 20
n.margin_bottom = 1
n.hz_align = 2
n.vt_align = 2
n.rotation = 270
n.inner_background.color = "white"
n.opacity = 1.
tss.aligned_header.add_face(n, 0)
col_add = 1
else:
col_add = 1
if header_list:
for col, header in enumerate(header_list):
n = TextFace('%s' % (header))
n.margin_top = 0
n.margin_right = 1
n.margin_left = 20
n.margin_bottom = 1
n.rotation = 270
n.hz_align = 2
n.vt_align = 2
n.inner_background.color = "white"
n.opacity = 1.
tss.aligned_header.add_face(n, col + col_add)
col_add += col + 1
if header_list3:
#print 'header_list 3!'
col_tmp = 0
for header in header_list3:
n = TextFace('%s' % (header))
n.margin_top = 1
n.margin_right = 1
n.margin_left = 20
n.margin_bottom = 1
n.rotation = 270
n.hz_align = 2
n.vt_align = 2
n.inner_background.color = "white"
n.opacity = 1.
if set2taxon2value_list_simple_barplot_counts:
if col_tmp == 0:
col_tmp += 1
tss.aligned_header.add_face(n, col_tmp + 1 + col_add)
n = TextFace(' ')
tss.aligned_header.add_face(n, col_tmp + col_add)
col_tmp += 2
else:
tss.aligned_header.add_face(n, col_tmp + col_add)
col_tmp += 1
if set2taxon2value_list_simple_barplot_counts:
col_add += col_tmp
else:
col_add += col_tmp
if header_list2:
for col, header in enumerate(header_list2):
n = TextFace('%s' % (header))
n.margin_top = 1
n.margin_right = 1
n.margin_left = 20
n.margin_bottom = 1
n.rotation = 270
n.hz_align = 2
n.vt_align = 2
n.inner_background.color = "white"
n.opacity = 1.
tss.aligned_header.add_face(n, col + col_add)
col_add += col + 1
if taxon2label:
try:
n = TextFace('%s' % taxon2label[lf.name])
except:
try:
n = TextFace('%s' % taxon2label[int(lf.name)])
except:
n = TextFace('-')
n.margin_top = 1
n.margin_right = 1
n.margin_left = 20
n.margin_bottom = 1
n.inner_background.color = "white"
n.opacity = 1.
if rotate:
n.rotation = 270
lf.add_face(n, 1, position="aligned")
col_add = 2
else:
col_add = 2
if taxon2value_list_barplot:
try:
val_list_of_lists = taxon2value_list_barplot[lf.name]
except:
val_list_of_lists = taxon2value_list_barplot[int(lf.name)]
#col_count = 0
for col, value_list in enumerate(val_list_of_lists):
total = float(sum(value_list))
percentages = [(i / total) * 100 for i in value_list]
if col % 3 == 0:
col_list = colors2
else:
col_list = colors
b = StackedBarFace(percentages,
width=150,
height=18,
colors=col_list[0:len(percentages)])
b.rotation = 0
b.inner_border.color = "white"
b.inner_border.width = 0
b.margin_right = 5
b.margin_left = 5
if rotate:
b.rotation = 270
lf.add_face(b, col + col_add, position="aligned")
#col_count+=1
col_add += col + 1
if set2taxon2value_list_simple_barplot:
col_list = [
'#fc8d59', '#91bfdb', '#99d594', '#c51b7d', '#f1a340',
'#999999'
]
color_i = 0
col = 0
for one_set in header_list3:
if color_i > 5:
color_i = 0
color = col_list[color_i]
color_i += 1
# values for all taxons
values_lists = [
float(i) for i in
set2taxon2value_list_simple_barplot[one_set].values()
]
#print values_lists
#print one_set
value = set2taxon2value_list_simple_barplot[one_set][lf.name]
if set2taxon2value_list_simple_barplot_counts:
if isinstance(value, float):
a = TextFace(" %s " % str(round(value, 2)))
else:
a = TextFace(" %s " % str(value))
a.margin_top = 1
a.margin_right = 2
a.margin_left = 5
a.margin_bottom = 1
if rotate:
a.rotation = 270
lf.add_face(a, col + col_add, position="aligned")
#print 'value and max', value, max(values_lists)
fraction_biggest = (float(value) / max(values_lists)) * 100
fraction_rest = 100 - fraction_biggest
#print 'fractions', fraction_biggest, fraction_rest
b = StackedBarFace([fraction_biggest, fraction_rest],
width=100,
height=15,
colors=[color, 'white'])
b.rotation = 0
b.inner_border.color = "grey"
b.inner_border.width = 0
b.margin_right = 15
b.margin_left = 0
if rotate:
b.rotation = 270
if set2taxon2value_list_simple_barplot_counts:
if col == 0:
col += 1
lf.add_face(b, col + 1 + col_add, position="aligned")
col += 2
else:
lf.add_face(b, col + col_add, position="aligned")
col += 1
if set2taxon2value_list_simple_barplot_counts:
col_add += col
else:
col_add += col
if taxon2set2value_heatmap:
i = 0
#if not taxon2label:
# col_add-=1
for col2, head in enumerate(header_list2):
col_name = header_list2[i]
try:
value = taxon2set2value_heatmap[col_name][str(lf.name)]
except:
try:
value = taxon2set2value_heatmap[col_name][round(
float(lf.name), 2)]
except:
value = 0
if header_list2[i] == 'duplicates':
print('dupli', lf.name, value)
#print 'val----------------', value
if int(value) > 0:
if int(value) >= 10 and int(value) < 100:
n = TextFace('%4i' % value)
elif int(value) >= 100:
n = TextFace('%3i' % value)
else:
n = TextFace('%5i' % value)
n.margin_top = 1
n.margin_right = 2
n.margin_left = 5
n.margin_bottom = 1
n.hz_align = 1
n.vt_align = 1
if rotate:
n.rotation = 270
n.inner_background.color = rgb2hex(
column2scale[col_name][0].to_rgba(
float(value))) #"orange"
#print 'xaxaxaxaxa', value,
if float(value) > column2scale[col_name][1]:
n.fgcolor = 'white'
n.opacity = 1.
n.hz_align = 1
n.vt_align = 1
lf.add_face(n, col2 + col_add, position="aligned")
i += 1
else:
n = TextFace('')
n.margin_top = 1
n.margin_right = 1
n.margin_left = 5
n.margin_bottom = 1
n.inner_background.color = "white"
n.opacity = 1.
if rotate:
n.rotation = 270
lf.add_face(n, col2 + col_add, position="aligned")
i += 1
#lf.name = taxon2description[lf.name]
n = TextFace(taxon2description[lf.name],
fgcolor="black",
fsize=12,
fstyle='italic')
lf.add_face(n, 0)
for n in t1.traverse():
nstyle = NodeStyle()
if n.support < 1:
nstyle["fgcolor"] = "black"
nstyle["size"] = 6
n.set_style(nstyle)
else:
nstyle["fgcolor"] = "red"
nstyle["size"] = 0
n.set_style(nstyle)
return t1, tss
def add_simple_barplot(self,
taxon2value,
header_name,
color=False,
show_values=False,
substract_min=False,
highlight_cutoff=False,
highlight_reverse=False,
max_value=False):
if not show_values:
self._add_header(header_name, column_add=0)
else:
self._add_header(header_name, column_add=1)
values_lists = [float(i) for i in taxon2value.values()]
min_value = min(values_lists)
if substract_min:
values_lists = [i-min_value for i in values_lists]
for taxon in list(taxon2value.keys()):
taxon2value[taxon] = taxon2value[taxon]-min_value
if not color:
color = self._get_default_barplot_color()
for i, lf in enumerate(self.tree.iter_leaves()):
try:
value = taxon2value[lf.name]
except KeyError:
value = 0
if show_values:
barplot_column = 1
if substract_min:
real_value = value + min_value
else:
real_value = value
if isinstance(real_value, float):
a = TextFace(" %s " % str(round(real_value,2)))
else:
a = TextFace(" %s " % str(real_value))
a.margin_top = 1
a.margin_right = 2
a.margin_left = 5
a.margin_bottom = 1
if self.rotate:
a.rotation = 270
lf.add_face(a, self.column_count, position="aligned")
else:
barplot_column = 0
if not max_value:
fraction_biggest = (float(value)/max(values_lists))*100
else:
fraction_biggest = (float(value)/max_value)*100
fraction_rest = 100-fraction_biggest
if highlight_cutoff:
if substract_min:
real_value = value + min_value
else:
real_value = value
if highlight_reverse:
if real_value > highlight_cutoff:
lcolor = "grey"
else:
lcolor = color
else:
if real_value < highlight_cutoff:
lcolor = "grey"
else:
lcolor = color
else:
lcolor = color
b = StackedBarFace([fraction_biggest, fraction_rest], width=100, height=15,colors=[lcolor, 'white'])
b.rotation= 0
b.inner_border.color = "grey"
b.inner_border.width = 0
b.margin_right = 15
b.margin_left = 0
if self.rotate:
b.rotation = 270
lf.add_face(b, self.column_count + barplot_column, position="aligned")
self.column_count += (1 + barplot_column)
def plot_ete_tree(tree_file,
ordered_queries,
leaf_id2protein_id2identity,
leaf_id2mlst,
leaf_id2spa,
leaf_id2meca,
show_identity_values=True,
leaf_id2description=False):
mlst_list = list(set(leaf_id2mlst.values()))
mlst2color = dict(zip(mlst_list, get_spaced_colors(len(mlst_list))))
mlst2color['-'] = 'white'
t1 = Tree(tree_file)
tss = TreeStyle()
R = t1.get_midpoint_outgroup()
t1.set_outgroup(R)
t1.ladderize()
head = True
column_add = 4
for lf in t1.iter_leaves():
lf.branch_vertical_margin = 0
# add MLST
if head:
n = TextFace(' MLST ')
n.margin_top = 2
n.margin_right = 2
n.margin_left = 2
n.margin_bottom = 2
n.rotation = 270
n.vt_align = 2
n.hz_align = 2
n.inner_background.color = "white"
n.opacity = 1.
tss.aligned_header.add_face(n, 1)
if lf.name in leaf2mlst:
n = TextFace(' %s ' % leaf_id2mlst[lf.name])
n.inner_background.color = 'white'
m = TextFace(' ')
m.inner_background.color = mlst2color[leaf_id2mlst[lf.name]]
else:
n = TextFace(' na ')
n.inner_background.color = "grey"
m = TextFace(' ')
m.inner_background.color = "white"
n.opacity = 1.
n.margin_top = 2
n.margin_right = 2
n.margin_left = 0
n.margin_bottom = 2
m.margin_top = 2
m.margin_right = 0
m.margin_left = 20
m.margin_bottom = 2
lf.add_face(m, 0, position="aligned")
lf.add_face(n, 1, position="aligned")
# add spa typing
if head:
n = TextFace(' spa ')
n.margin_top = 2
n.margin_right = 2
n.margin_left = 2
n.margin_bottom = 2
n.rotation = 270
n.vt_align = 2
n.hz_align = 2
n.inner_background.color = "white"
n.opacity = 1.
tss.aligned_header.add_face(n, column_add-2)
if lf.name in leaf_id2spa:
n = TextFace(' %s ' % leaf_id2spa[lf.name])
n.inner_background.color = "white"
else:
n = TextFace(' na ')
n.inner_background.color = "grey"
n.opacity = 1.
n.margin_top = 2
n.margin_right = 2
n.margin_left = 2
n.margin_bottom = 2
lf.add_face(n, column_add-2, position="aligned")
# add mecA typing
if head:
n = TextFace(' mecA ')
n.margin_top = 2
n.margin_right = 2
n.margin_left = 2
n.margin_bottom = 2
n.rotation = 270
n.vt_align = 2
n.hz_align = 2
n.inner_background.color = "white"
n.opacity = 1.
tss.aligned_header.add_face(n, column_add-1)
if lf.name in leaf_id2meca:
n = TextFace(' %s ' % leaf_id2meca[lf.name])
if leaf_id2meca[lf.name] == 'Perfect':
n.inner_background.color = "red"
elif leaf_id2meca[lf.name] == 'Strict':
n.inner_background.color = "orange"
else:
n.inner_background.color = "white"
else:
n = TextFace(' na ')
n.inner_background.color = "grey"
n.opacity = 1.
n.margin_top = 2
n.margin_right = 2
n.margin_left = 2
n.margin_bottom = 2
lf.add_face(n, column_add-1, position="aligned")
# loop to add virulence gene hits
for column, protein_id in enumerate(ordered_queries):
# draw labels at the top of each column
if head:
if show_identity_values:
n = TextFace(' %s ' % str(protein_id))
n.margin_top = 2
n.margin_right = 2
n.margin_left = 2
n.margin_bottom = 2
n.rotation = 270
n.vt_align = 2
n.hz_align = 2
n.inner_background.color = "white"
n.opacity = 1.
tss.aligned_header.add_face(n, column+column_add)
else:
n = TextFace(' %s ' % str(protein_id), fsize=6)
n.margin_top = 0
n.margin_right = 0
n.margin_left = 0
n.margin_bottom = 0
n.rotation = 270
n.vt_align = 2
n.hz_align = 2
n.inner_background.color = "white"
n.opacity = 1.
# lf.add_face(n, col, position="aligned")
tss.aligned_header.add_face(n, column+column_add)
# draw column content
if lf.name not in leaf_id2protein_id2identity:
n = TextFace(' %s ' % str(' na '))
n.opacity = 1.
n.margin_top = 2
n.margin_right = 2
n.margin_left = 2
n.margin_bottom = 2
n.inner_background.color = "grey"
lf.add_face(n, column+column_add, position="aligned")
else:
if protein_id in leaf_id2protein_id2identity[lf.name]:
identity_value = float(leaf_id2protein_id2identity[lf.name][protein_id])
color = rgb2hex(m_blue.to_rgba(identity_value))
if show_identity_values:
# report identity values in coloured boxes
# adapt box size depending the digit width
if str(identity_value) == '100.00' or str(identity_value) == '100.0':
identity_value = '100'
n = TextFace(" %s " % identity_value)
else:
n = TextFace("%.2f" % round(float(identity_value), 2))
# color text to white for dark cells
if float(identity_value) > 95:
n.fgcolor = "white"
n.opacity = 1.
n.margin_top = 2
n.margin_right = 2
n.margin_left = 2
n.margin_bottom = 2
n.inner_background.color = color
lf.add_face(n, column+column_add, position="aligned")
else:
# draw coloured boxes without text
n = TextFace(' ')
n.margin_top = 0
n.margin_right = 0
n.margin_left = 0
n.margin_bottom = 0
# n.color = color
n.inner_background.color = color
lf.add_face(n, column+column_add, position="aligned")
else:
n = TextFace(' %s ' % str(' - '))
n.opacity = 1.
n.margin_top = 2
n.margin_right = 2
n.margin_left = 2
n.margin_bottom = 2
n.inner_background.color = "white"
lf.add_face(n, column+column_add, position="aligned")
# end of first leaf: turn off header
head = False
# add boostrap supports
for n in t1.traverse():
nstyle = NodeStyle()
if n.support < 0.9:
nstyle["fgcolor"] = "blue"
nstyle["size"] = 6
n.set_style(nstyle)
else:
nstyle["fgcolor"] = "red"
nstyle["size"] = 0
n.set_style(nstyle)
return t1, tss
def heatmap_view(tree, orthologous_groups, save_dir):
"""Generates a heatmap of regulation states in all species."""
light_tree = copy.deepcopy(tree) # Tree copy for the light heatmap
# Heat map settings
rect_face_fgcolor = 'black'
locus_tag_len = max(
len(gene.locus_tag) + 5 for ortho_grp in orthologous_groups
for gene in ortho_grp.genes)
rect_face_width = locus_tag_len * 8
light_rect_face_width = 20
rect_face_height = 20
rotation = 90
# Sort orthologous groups by the number of regulated genes in each group
orthologous_groups = filter_and_sort_orthologous_grps(orthologous_groups)
# For each species and its gene in each orthologous group, draw a rectangle
for node, light_node in zip(tree.get_leaves(), light_tree.get_leaves()):
for i, orthologous_grp in enumerate(orthologous_groups, start=1):
#get all orthologs in group
matching_genes = [g for g in orthologous_grp.genes \
if g.genome.strain_name == node.name]
#if there is ortholog
if len(matching_genes) > 0:
# Get the first ortholog from the genome in the group
#this is the one with higher probability of regulation.
#so this probability will be displayed for the group
gene = matching_genes[0]
p_regulation = gene.operon.regulation_probability
p_notregulation = 1.0 - p_regulation
p_absence = 0
# No ortholog from this genome
else:
gene = None
p_regulation = 0
p_notregulation = 0
p_absence = 1
# Color of the rectangle is based on probabilities
rect_face_bgcolor = rgb2hex(p_notregulation, p_regulation,
p_absence)
rect_face_text = ('%s [%d]' %
(gene.locus_tag, gene.operon.operon_id)
if gene else '')
rect_face_label = {
'text': rect_face_text,
'font': 'Courier',
'fontsize': 8,
'color': 'black'
}
# Create the rectangle
rect_face = RectFace(rect_face_width,
rect_face_height,
rect_face_fgcolor,
rect_face_bgcolor,
label=rect_face_label)
light_rect_face = RectFace(light_rect_face_width,
rect_face_height,
rect_face_fgcolor,
rect_face_bgcolor,
label='')
rect_face.rotation = -rotation
light_rect_face.rotation = -rotation
# Add the rectangle to the corresponding column
node.add_face(rect_face, column=i, position='aligned')
light_node.add_face(light_rect_face, column=i, position='aligned')
ts = TreeStyle()
# Add orthologous group descriptions
descriptions = ['-'.join([grp.description, \
str([item['ID'] for item in grp.COGs]) if len(grp.COGs)>0 else '', \
str([item['ID'] for item in grp.NOGs]) if len(grp.NOGs)>0 else '', \
str([item['ID'] for item in grp.PFAMs])] if len(grp.PFAMs)>0 else '')\
for grp in orthologous_groups]
max_description_len = max(map(len, descriptions))
descriptions = [
'[%d]' % i + description + ' ' *
(max_description_len - len(description))
for i, description in enumerate(descriptions, start=1)
]
for i, description in enumerate(descriptions, start=1):
text_face = TextFace(description, ftype='Courier')
text_face.hz_align = 1
text_face.vt_align = 1
text_face.rotation = -rotation
ts.aligned_header.add_face(text_face, column=i)
# Rotate the generated heatmap.
ts.margin_left = 10
ts.margin_top = 20
ts.rotation = rotation
ts.show_scale = False
# For some reason, it can't render to PDF in color
tree.render(os.path.join(save_dir, 'heatmap.svg'), tree_style=ts)
light_tree.render(os.path.join(save_dir, 'heatmap_light.svg'),
tree_style=ts)
def draw_tree(variant_dict, profile_dict, the_tree, ref_name):
major_alleles, minor_alleles = profile_dict
t = Tree(the_tree, quoted_node_names=True)
font_size = 8
font_type = 'Heveltica'
font_gap = 3
font_buffer = 10
ts = TreeStyle()
position_list = None
allele_count = {}
max_major = 0
max_minor = 0
max_combo = 0
max_major_minor = 0
for n in t.iter_leaves():
the_name = n.name
if the_name == ref_name:
t.set_outgroup(n)
if position_list is None:
position_list = list(variant_dict[the_name])
position_list.sort()
for num, i in enumerate(position_list):
nameF = TextFace(font_gap * ' ' + str(i) + ' ' * font_buffer,
fsize=font_size,
ftype=font_type,
tight_text=True)
nameF.rotation = -90
ts.aligned_header.add_face(nameF, column=num)
minor_allele, major_allele, missing = 0, 0, 0
for num, i in enumerate(position_list):
if i in major_alleles and variant_dict[the_name][
i] in major_alleles[i]:
s, l = 0.0, 0.1
major_allele += 1
elif i in minor_alleles and variant_dict[the_name][
i] in minor_alleles[i]:
s, l = 0.0, 0.5
minor_allele += 1
else:
s, l = 0.0, 0.9
missing += 1
if variant_dict[the_name][i] == 'a':
colour = colorstr(hsl_to_rgb(0, s, l))
#colour = "#65dad0"
elif variant_dict[the_name][i] == 't':
colour = colorstr(hsl_to_rgb(140, s, l))
#colour = "#daa3dc"
elif variant_dict[the_name][i] == 'c':
colour = colorstr(hsl_to_rgb(220, s, l))
#colour = "#9bd686"
elif variant_dict[the_name][i] == 'g':
colour = colorstr(hsl_to_rgb(300, s, l))
#colour = "#e1b86f"
else:
colour = "#ffffff"
#colour = colorstr(hsl_to_rgb(h, s, l))
n.add_face(RectFace(20, 20, colour, colour),
column=num,
position="aligned")
allele_count[the_name] = (major_allele, minor_allele, missing)
if major_allele > max_major:
max_major = major_allele
if minor_allele + major_allele >= max_combo:
max_combo = minor_allele + major_allele
if minor_allele >= max_minor:
max_minor = minor_allele
if major_allele > max_major_minor:
max_major_minor = major_allele
nst1 = NodeStyle()
nst1["bgcolor"] = "LightSteelBlue"
nst2 = NodeStyle()
nst2["bgcolor"] = "Moccasin"
with open(args.out_text, 'w') as o:
for n in t.iter_leaves():
the_name = n.name
major_allele, minor_allele, missing = allele_count[the_name]
if major_allele == max_major:
o.write("max_major\t%s\t%d\t%d\t%d\n" %
(the_name, major_allele, minor_allele, missing))
n.img_style["bgcolor"] = "#cb6a49"
elif minor_allele == max_minor and major_allele == max_major_minor:
o.write("max_minor\t%s\t%d\t%d\t%d\n" %
(the_name, major_allele, minor_allele, missing))
n.img_style["bgcolor"] = "#7aa457"
elif minor_allele + major_allele == max_combo:
o.write("max_combo\t%s\t%d\t%d\t%d\n" %
(the_name, major_allele, minor_allele, missing))
n.img_style["bgcolor"] = "#a46cb7"
n.add_face(TextFace('%d/%d/%d' %
(major_allele, minor_allele, missing),
fsize=font_size,
ftype=font_type,
tight_text=True),
column=len(position_list),
position="aligned")
ts.legend.add_face(TextFace('A',
fsize=font_size,
ftype=font_type,
tight_text=True),
column=0)
s = 0.5
ts.legend.add_face(RectFace(20, 20, colorstr(hsl_to_rgb(0, s, 0.3)),
colorstr(hsl_to_rgb(0, s, 0.3))),
column=1)
ts.legend.add_face(RectFace(20, 20, colorstr(hsl_to_rgb(0, s, 0.5)),
colorstr(hsl_to_rgb(0, s, 0.5))),
column=2)
ts.legend.add_face(RectFace(20, 20, colorstr(hsl_to_rgb(0, s, 0.8)),
colorstr(hsl_to_rgb(0, s, 0.8))),
column=3)
ts.legend.add_face(TextFace('T',
fsize=font_size,
ftype=font_type,
tight_text=True),
column=0)
ts.legend.add_face(RectFace(20, 20, colorstr(hsl_to_rgb(140, s, 0.3)),
colorstr(hsl_to_rgb(140, s, 0.3))),
column=1)
ts.legend.add_face(RectFace(20, 20, colorstr(hsl_to_rgb(140, s, 0.5)),
colorstr(hsl_to_rgb(140, s, 0.5))),
column=2)
ts.legend.add_face(RectFace(20, 20, colorstr(hsl_to_rgb(140, s, 0.8)),
colorstr(hsl_to_rgb(140, s, 0.8))),
column=3)
ts.legend.add_face(TextFace('C',
fsize=font_size,
ftype=font_type,
tight_text=True),
column=0)
ts.legend.add_face(RectFace(20, 20, colorstr(hsl_to_rgb(220, s, 0.3)),
colorstr(hsl_to_rgb(220, s, 0.3))),
column=1)
ts.legend.add_face(RectFace(20, 20, colorstr(hsl_to_rgb(220, s, 0.5)),
colorstr(hsl_to_rgb(220, s, 0.5))),
column=2)
ts.legend.add_face(RectFace(20, 20, colorstr(hsl_to_rgb(220, s, 0.8)),
colorstr(hsl_to_rgb(220, s, 0.8))),
column=3)
ts.legend.add_face(TextFace('G',
fsize=font_size,
ftype=font_type,
tight_text=True),
column=0)
ts.legend.add_face(RectFace(20, 20, colorstr(hsl_to_rgb(300, s, 0.3)),
colorstr(hsl_to_rgb(300, s, 0.3))),
column=1)
ts.legend.add_face(RectFace(20, 20, colorstr(hsl_to_rgb(300, s, 0.5)),
colorstr(hsl_to_rgb(300, s, 0.5))),
column=2)
ts.legend.add_face(RectFace(20, 20, colorstr(hsl_to_rgb(300, s, 0.8)),
colorstr(hsl_to_rgb(300, s, 0.8))),
column=3)
ts.legend.add_face(TextFace('-',
fsize=font_size,
ftype=font_type,
tight_text=True),
column=0)
ts.legend.add_face(RectFace(20, 20, "#cccccc", "#cccccc"), column=1)
t.render(args.out_file, w=210, units='mm', tree_style=ts)
def plot_phylum_counts(NOG_id,
rank='phylum',
colapse_low_species_counts=4,
remove_unlassified=True):
'''
1. get phylum tree
2. foreach species => get phylum
3. build phylum2count dictionnary
3. plot barchart
# merge eukaryotes into 5 main clades
# merge virus as a single clade
ATTENTION: no-rank groups and no-rank species...
'''
import MySQLdb
import os
from chlamdb.biosqldb import manipulate_biosqldb
from ete3 import NCBITaxa, Tree, TextFace, TreeStyle, StackedBarFace
ncbi = NCBITaxa()
sqlpsw = os.environ['SQLPSW']
conn = MySQLdb.connect(
host="localhost", # your host, usually localhost
user="******", # your username
passwd=sqlpsw, # your password
db="eggnog") # name of the data base
cursor = conn.cursor()
sql = 'select * from eggnog.leaf2n_genomes_%s' % rank
cursor.execute(sql, )
leaf_taxon2n_species = manipulate_biosqldb.to_dict(cursor.fetchall())
leaf_taxon2n_species_with_domain = get_NOG_taxonomy(NOG_id, rank)
sql = 'select phylogeny from eggnog.phylogeny where rank="%s"' % (rank)
cursor.execute(sql, )
tree = Tree(cursor.fetchall()[0][0], format=1)
sql = 'select * from eggnog.taxid2label_%s' % rank
cursor.execute(sql, )
taxon_id2scientific_name_and_rank = manipulate_biosqldb.to_dict(
cursor.fetchall())
taxon_id2scientific_name_and_rank = {
str(k): v
for k, v in taxon_id2scientific_name_and_rank.items()
}
tss = TreeStyle()
tss.draw_guiding_lines = True
tss.guiding_lines_color = "blue"
keep = []
for lf in tree.iter_leaves():
# n genomes
if remove_unlassified:
label = taxon_id2scientific_name_and_rank[str(lf.name)][0]
if 'unclassified' in label:
continue
n_genomes = int(leaf_taxon2n_species[lf.name])
if n_genomes > colapse_low_species_counts:
keep.append(lf.name)
print('number of leaaves:', len(keep))
tree.prune(keep)
header_list = ['Rank', 'N genomes', 'N with %s' % NOG_id, 'Percentage']
for col, header in enumerate(header_list):
n = TextFace('%s' % (header))
n.margin_top = 0
n.margin_right = 1
n.margin_left = 20
n.margin_bottom = 1
n.rotation = 270
n.hz_align = 2
n.vt_align = 2
n.inner_background.color = "white"
n.opacity = 1.
tss.aligned_header.add_face(n, col)
for lf in tree.iter_leaves():
# n genomes
n_genomes = int(leaf_taxon2n_species[lf.name])
if n_genomes <= colapse_low_species_counts:
continue
n = TextFace(' %s ' % str(leaf_taxon2n_species[lf.name]))
n.margin_top = 1
n.margin_right = 1
n.margin_left = 0
n.margin_bottom = 1
n.fsize = 7
n.inner_background.color = "white"
n.opacity = 1.
lf.add_face(n, 2, position="aligned")
# n genomes with domain
try:
m = TextFace(' %s ' %
str(leaf_taxon2n_species_with_domain[lf.name]))
except:
m = TextFace(' 0 ')
m.margin_top = 1
m.margin_right = 1
m.margin_left = 0
m.margin_bottom = 1
m.fsize = 7
m.inner_background.color = "white"
m.opacity = 1.
lf.add_face(m, 3, position="aligned")
# rank
ranks = ncbi.get_rank([lf.name])
try:
r = ranks[max(ranks.keys())]
except:
r = '-'
n = TextFace(' %s ' % r, fsize=14, fgcolor='red')
n.margin_top = 1
n.margin_right = 1
n.margin_left = 0
n.margin_bottom = 1
n.fsize = 7
n.inner_background.color = "white"
n.opacity = 1.
lf.add_face(n, 1, position="aligned")
# percent with target domain
try:
percentage = (float(leaf_taxon2n_species_with_domain[lf.name]) /
float(leaf_taxon2n_species[lf.name])) * 100
except:
percentage = 0
m = TextFace(' %s ' % str(round(percentage, 2)))
m.fsize = 1
m.margin_top = 1
m.margin_right = 1
m.margin_left = 0
m.margin_bottom = 1
m.fsize = 7
m.inner_background.color = "white"
m.opacity = 1.
lf.add_face(m, 4, position="aligned")
b = StackedBarFace([percentage, 100 - percentage],
width=100,
height=10,
colors=["#7fc97f", "white"])
b.rotation = 0
b.inner_border.color = "grey"
b.inner_border.width = 0
b.margin_right = 15
b.margin_left = 0
lf.add_face(b, 5, position="aligned")
n = TextFace('%s' % taxon_id2scientific_name_and_rank[str(lf.name)][0],
fgcolor="black",
fsize=9) # , fstyle = 'italic'
lf.name = " %s (%s)" % (taxon_id2scientific_name_and_rank[str(
lf.name)][0], str(lf.name))
n.margin_right = 10
lf.add_face(n, 0)
tss.show_leaf_name = False
for node in tree.traverse("postorder"):
try:
r = taxon_id2scientific_name_and_rank[str(node.name)][1]
except:
pass
try:
if r in ['phylum', 'superkingdom', 'class', 'subphylum'
] or taxon_id2scientific_name_and_rank[str(
node.name)][0] in ['FCB group']:
hola = TextFace(
"%s" %
(taxon_id2scientific_name_and_rank[str(node.name)][0]))
node.add_face(hola, column=0, position="branch-top")
except:
pass
return tree, tss
def draw_tree(the_tree, colour, back_color, label, out_file, the_scale, extend,
bootstrap, group_file, grid_options, the_table, pres_abs,
circular):
t = Tree(the_tree, quoted_node_names=True)
# t.ladderize()
font_size = 8
font_type = 'Heveltica'
font_gap = 3
font_buffer = 10
o = t.get_midpoint_outgroup()
t.set_outgroup(o)
the_leaves = []
for leaves in t.iter_leaves():
the_leaves.append(leaves)
groups = {}
num = 0
# set cutoff value for clades as 1/20th of the distance between the furthest two branches
# assign nodes to groups
last_node = None
ca_list = []
if not group_file is None:
style = NodeStyle()
style['size'] = 0
style["vt_line_color"] = '#000000'
style["hz_line_color"] = '#000000'
style["vt_line_width"] = 1
style["hz_line_width"] = 1
for n in t.traverse():
n.set_style(style)
with open(group_file) as f:
group_dict = {}
for line in f:
group_dict[line.split()[0]] = line.split()[1]
for node in the_leaves:
i = node.name
for j in group_dict:
if j in i:
if group_dict[j] in groups:
groups[group_dict[j]].append(i)
else:
groups[group_dict[j]] = [i]
coloured_nodes = []
for i in groups:
the_col = i
style = NodeStyle()
style['size'] = 0
style["vt_line_color"] = the_col
style["hz_line_color"] = the_col
style["vt_line_width"] = 2
style["hz_line_width"] = 2
if len(groups[i]) == 1:
ca = t.search_nodes(name=groups[i][0])[0]
ca.set_style(style)
coloured_nodes.append(ca)
else:
ca = t.get_common_ancestor(groups[i])
ca.set_style(style)
coloured_nodes.append(ca)
tocolor = []
for j in ca.children:
tocolor.append(j)
while len(tocolor) > 0:
x = tocolor.pop(0)
coloured_nodes.append(x)
x.set_style(style)
for j in x.children:
tocolor.append(j)
ca_list.append((ca, the_col))
if back_color:
# for each common ancestor node get it's closest common ancestor neighbour and find the common ancestor of those two nodes
# colour the common ancestor then add it to the group - continue until only the root node is left
while len(ca_list) > 1:
distance = float('inf')
for i, col1 in ca_list:
for j, col2 in ca_list:
if not i is j:
parent = t.get_common_ancestor(i, j)
getit = True
the_dist = t.get_distance(i, j)
if the_dist <= distance:
distance = the_dist
the_i = i
the_j = j
the_i_col = col1
the_j_col = col2
ca_list.remove((the_i, the_i_col))
ca_list.remove((the_j, the_j_col))
rgb1 = strtorgb(the_i_col)
rgb2 = strtorgb(the_j_col)
rgb3 = ((rgb1[0] + rgb2[0]) / 2, (rgb1[1] + rgb2[1]) / 2,
(rgb1[2] + rgb2[2]) / 2)
new_col = colorstr(rgb3)
new_node = t.get_common_ancestor(the_i, the_j)
the_col = new_col
style = NodeStyle()
style['size'] = 0
style["vt_line_color"] = the_col
style["hz_line_color"] = the_col
style["vt_line_width"] = 2
style["hz_line_width"] = 2
new_node.set_style(style)
coloured_nodes.append(new_node)
ca_list.append((new_node, new_col))
for j in new_node.children:
tocolor.append(j)
while len(tocolor) > 0:
x = tocolor.pop(0)
if not x in coloured_nodes:
coloured_nodes.append(x)
x.set_style(style)
for j in x.children:
tocolor.append(j)
elif colour:
distances = []
for node1 in the_leaves:
for node2 in the_leaves:
if node1 != node2:
distances.append(t.get_distance(node1, node2))
distances.sort()
clade_cutoff = distances[len(distances) / 4]
for node in the_leaves:
i = node.name
if not last_node is None:
if t.get_distance(node, last_node) <= clade_cutoff:
groups[group_num].append(i)
else:
groups[num] = [num, i]
group_num = num
num += 1
else:
groups[num] = [num, i]
group_num = num
num += 1
last_node = node
for i in groups:
num = groups[i][0]
h = num * 360 / len(groups)
the_col = hsl_to_str(h, 0.5, 0.5)
style = NodeStyle()
style['size'] = 0
style["vt_line_color"] = the_col
style["hz_line_color"] = the_col
style["vt_line_width"] = 2
style["hz_line_width"] = 2
if len(groups[i]) == 2:
ca = t.search_nodes(name=groups[i][1])[0]
ca.set_style(style)
else:
ca = t.get_common_ancestor(groups[i][1:])
ca.set_style(style)
tocolor = []
for j in ca.children:
tocolor.append(j)
while len(tocolor) > 0:
x = tocolor.pop(0)
x.set_style(style)
for j in x.children:
tocolor.append(j)
ca_list.append((ca, h))
# for each common ancestor node get it's closest common ancestor neighbour and find the common ancestor of those two nodes
# colour the common ancestor then add it to the group - continue until only the root node is left
while len(ca_list) > 1:
distance = float('inf')
got_one = False
for i, col1 in ca_list:
for j, col2 in ca_list:
if not i is j:
parent = t.get_common_ancestor(i, j)
getit = True
for children in parent.children:
if children != i and children != j:
getit = False
break
if getit:
the_dist = t.get_distance(i, j)
if the_dist <= distance:
distance = the_dist
the_i = i
the_j = j
the_i_col = col1
the_j_col = col2
got_one = True
if not got_one:
break
ca_list.remove((the_i, the_i_col))
ca_list.remove((the_j, the_j_col))
new_col = (the_i_col + the_j_col) / 2
new_node = t.get_common_ancestor(the_i, the_j)
the_col = hsl_to_str(new_col, 0.5, 0.3)
style = NodeStyle()
style['size'] = 0
style["vt_line_color"] = the_col
style["hz_line_color"] = the_col
style["vt_line_width"] = 2
style["hz_line_width"] = 2
new_node.set_style(style)
ca_list.append((new_node, new_col))
# if you just want a black tree
else:
style = NodeStyle()
style['size'] = 0
style["vt_line_color"] = '#000000'
style["hz_line_color"] = '#000000'
style["vt_line_width"] = 1
style["hz_line_width"] = 1
for n in t.traverse():
n.set_style(style)
color_list = [(240, 163, 255), (0, 117, 220), (153, 63, 0), (76, 0, 92),
(25, 25, 25), (0, 92, 49), (43, 206, 72), (255, 204, 153),
(128, 128, 128), (148, 255, 181), (143, 124, 0),
(157, 204, 0), (194, 0, 136), (0, 51, 128), (255, 164, 5),
(255, 168, 187), (66, 102, 0), (255, 0, 16), (94, 241, 242),
(0, 153, 143), (224, 255, 102), (116, 10, 255), (153, 0, 0),
(255, 255, 128), (255, 255, 0), (255, 80, 5), (0, 0, 0),
(50, 50, 50)]
up_to_colour = {}
ts = TreeStyle()
column_list = []
width_dict = {}
if not grid_options is None:
colour_dict = {}
type_dict = {}
min_val_dict = {}
max_val_dict = {}
leaf_name_dict = {}
header_count = 0
the_columns = {}
if grid_options == 'auto':
with open(the_table) as f:
headers = f.readline().rstrip().split('\t')[1:]
for i in headers:
the_columns[i] = [i]
type_dict[i] = 'colour'
colour_dict[i] = {'empty': '#FFFFFF'}
width_dict[i] = 20
up_to_colour[i] = 0
column_list.append(i)
else:
with open(grid_options) as g:
for line in g:
if line.startswith('H'):
name, type, width = line.rstrip().split('\t')[1:]
if name in the_columns:
the_columns[name].append(name + '_' +
str(header_count))
else:
the_columns[name] = [
name + '_' + str(header_count)
]
width = int(width)
name = name + '_' + str(header_count)
header_count += 1
colour_dict[name] = {'empty': '#FFFFFF'}
type_dict[name] = type
width_dict[name] = width
column_list.append(name)
up_to_colour[name] = 0
min_val_dict[name] = float('inf')
max_val_dict[name] = 0
elif line.startswith('C'):
c_name, c_col = line.rstrip().split('\t')[1:]
if not c_col.startswith('#'):
c_col = colorstr(map(int, c_col.split(',')))
colour_dict[name][c_name] = c_col
val_dict = {}
with open(the_table) as f:
headers = f.readline().rstrip().split('\t')[1:]
column_no = {}
for num, i in enumerate(headers):
if i in the_columns:
column_no[num] = i
for line in f:
name = line.split('\t')[0]
leaf_name = None
for n in t.traverse():
if n.is_leaf():
if name.split('.')[0] in n.name:
leaf_name = n.name
if leaf_name is None:
continue
else:
leaf_name_dict[leaf_name] = name
vals = line.rstrip().split('\t')[1:]
if name in val_dict:
sys.exit('Duplicate entry found in table.')
else:
val_dict[name] = {}
for num, val in enumerate(vals):
if num in column_no and val != '':
for q in the_columns[column_no[num]]:
column_name = q
if type_dict[column_name] == 'colour':
val_dict[name][column_name] = val
if not val in colour_dict[column_name]:
colour_dict[column_name][val] = colorstr(
color_list[up_to_colour[column_name] %
len(color_list)])
up_to_colour[column_name] += 1
elif type_dict[column_name] == 'text':
val_dict[name][column_name] = val
elif type_dict[column_name] == 'colour_scale_date':
year, month, day = val.split('-')
year, month, day = int(year), int(month), int(
day)
the_val = datetime.datetime(
year, month, day, 0, 0,
0) - datetime.datetime(
1970, 1, 1, 0, 0, 0)
val_dict[name][
column_name] = the_val.total_seconds()
if the_val.total_seconds(
) < min_val_dict[column_name]:
min_val_dict[
column_name] = the_val.total_seconds()
if the_val.total_seconds(
) > max_val_dict[column_name]:
max_val_dict[
column_name] = the_val.total_seconds()
elif type_dict[column_name] == 'colour_scale':
the_val = float(val)
val_dict[name][column_name] = the_val
if the_val < min_val_dict[column_name]:
min_val_dict[column_name] = the_val
if the_val > max_val_dict[column_name]:
max_val_dict[column_name] = the_val
else:
sys.exit('Unknown column type')
if not out_file is None:
new_desc = open(out_file + '.new_desc', 'w')
else:
new_desc = open('viridis.new_desc', 'w')
ts.legend_position = 3
leg_column = 0
for num, i in enumerate(column_list):
nameF = TextFace(font_gap * ' ' + i.rsplit('_', 1)[0] +
' ' * font_buffer,
fsize=font_size,
ftype=font_type,
tight_text=True)
nameF.rotation = -90
ts.aligned_header.add_face(nameF, column=num + 1)
new_desc.write('H\t' + i.rsplit('_', 1)[0] + '\t' + type_dict[i] +
'\t' + str(width_dict[i]) + '\n')
x = num * 200
if type_dict[i] == 'colour':
ts.legend.add_face(TextFace(
font_gap * ' ' + i.rsplit('_', 1)[0] + ' ' * font_buffer,
fsize=font_size,
ftype=font_type,
tight_text=True),
column=leg_column + 1)
ts.legend.add_face(RectFace(width_dict[i], 20, '#FFFFFF',
'#FFFFFF'),
column=leg_column)
for num2, j in enumerate(colour_dict[i]):
new_desc.write('C\t' + j + '\t' + colour_dict[i][j] + '\n')
ts.legend.add_face(TextFace(font_gap * ' ' + j +
' ' * font_buffer,
fsize=font_size,
ftype=font_type,
tight_text=True),
column=leg_column + 1)
ts.legend.add_face(RectFace(width_dict[i], 20,
colour_dict[i][j],
colour_dict[i][j]),
column=leg_column)
leg_column += 2
elif type_dict[i] == 'colour_scale':
ts.legend.add_face(TextFace(
font_gap * ' ' + i.rsplit('_', 1)[0] + ' ' * font_buffer,
fsize=font_size,
ftype=font_type,
tight_text=True),
column=leg_column + 1)
ts.legend.add_face(RectFace(width_dict[i], 20, '#FFFFFF',
'#FFFFFF'),
column=leg_column)
for num2 in range(11):
y = num2 * 20 + 30
val = (max_val_dict[i] - min_val_dict[i]) * num2 / 10.0
h = val / (max_val_dict[i] - min_val_dict[i]) * 270
s = 0.5
l = 0.5
colour = hsl_to_str(h, s, l)
ts.legend.add_face(TextFace(font_gap * ' ' + str(val) +
' ' * font_buffer,
fsize=font_size,
ftype=font_type,
tight_text=True),
column=leg_column + 1)
ts.legend.add_face(RectFace(width_dict[i], 20, colour,
colour),
column=leg_column)
leg_column += 2
elif type_dict[i] == 'colour_scale_date':
ts.legend.add_face(TextFace(
font_gap * ' ' + i.rsplit('_', 1)[0] + ' ' * font_buffer,
fsize=font_size,
ftype=font_type,
tight_text=True),
column=leg_column + 1)
ts.legend.add_face(RectFace(width_dict[i], 20, '#FFFFFF',
'#FFFFFF'),
column=leg_column)
for num2 in range(11):
y = num2 * 20 + 30
val = (max_val_dict[i] - min_val_dict[i]) * num2 / 10.0
h = val / (max_val_dict[i] - min_val_dict[i]) * 360
s = 0.5
l = 0.5
colour = hsl_to_str(h, s, l)
days = str(int(val / 60 / 60 / 24)) + ' days'
ts.legend.add_face(TextFace(font_gap * ' ' + days +
' ' * font_buffer,
fsize=font_size,
ftype=font_type,
tight_text=True),
column=leg_column + 1)
ts.legend.add_face(RectFace(width_dict[i], 20, colour,
colour),
column=leg_column)
leg_column += 2
for n in t.traverse():
if n.is_leaf():
name = leaf_name_dict[n.name]
if i in val_dict[name]:
val = val_dict[name][i]
else:
val = 'empty'
if type_dict[i] == 'colour':
n.add_face(RectFace(width_dict[i], 20,
colour_dict[i][val],
colour_dict[i][val]),
column=num + 1,
position="aligned")
elif type_dict[i] == 'colour_scale' or type_dict[
i] == 'colour_scale_date':
if val == 'empty':
colour = '#FFFFFF'
else:
h = (val - min_val_dict[i]) / (
max_val_dict[i] - min_val_dict[i]) * 360
s = 0.5
l = 0.5
colour = hsl_to_str(h, s, l)
n.add_face(RectFace(width_dict[i], 20, colour, colour),
column=num + 1,
position="aligned")
elif type_dict[i] == 'text':
n.add_face(TextFace(font_gap * ' ' + val +
' ' * font_buffer,
fsize=font_size,
ftype=font_type,
tight_text=True),
column=num + 1,
position="aligned")
if not pres_abs is None:
starting_col = len(column_list) + 1
subprocess.Popen('makeblastdb -out tempdb -dbtype prot -in ' +
pres_abs[0],
shell=True).wait()
folder = pres_abs[1]
len_dict = {}
gene_list = []
ts.legend.add_face(TextFace(font_gap * ' ' + 'Gene present/absent' +
' ' * font_buffer,
fsize=font_size,
ftype=font_type,
tight_text=True),
column=starting_col + 1)
ts.legend.add_face(RectFace(20, 20, '#FFFFFF', '#FFFFFF'),
column=starting_col)
ts.legend.add_face(TextFace(font_gap * ' ' + 'Gene present/absent' +
' ' * font_buffer,
fsize=font_size,
ftype=font_type,
tight_text=True),
column=starting_col + 1)
ts.legend.add_face(RectFace(20, 20, "#5ba965", "#5ba965"),
column=starting_col)
ts.legend.add_face(TextFace(font_gap * ' ' + 'Gene present/absent' +
' ' * font_buffer,
fsize=font_size,
ftype=font_type,
tight_text=True),
column=starting_col + 1)
ts.legend.add_face(RectFace(20, 20, "#cb5b4c", "#cb5b4c"),
column=starting_col)
with open(pres_abs[0]) as f:
for line in f:
if line.startswith('>'):
name = line.split()[0][1:]
gene_list.append(name)
len_dict[name] = 0
nameF = TextFace(font_gap * ' ' + name + ' ' * font_buffer,
fsize=font_size,
ftype=font_type,
tight_text=True)
nameF.rotation = -90
ts.aligned_header.add_face(nameF,
column=starting_col +
len(gene_list) - 1)
else:
len_dict[name] += len(line.rstrip())
min_length = 0.9
min_ident = 90
for n in t.iter_leaves():
the_name = n.name
if the_name[0] == '"' and the_name[-1] == '"':
the_name = the_name[1:-1]
if the_name.endswith('.ref'):
the_name = the_name[:-4]
if not os.path.exists(folder + '/' + the_name):
for q in os.listdir(folder):
if q.startswith(the_name):
the_name = q
if not os.path.exists(the_name + '.blast'):
subprocess.Popen(
'blastx -query ' + folder + '/' + the_name +
' -db tempdb -outfmt 6 -num_threads 24 -out ' + the_name +
'.blast',
shell=True).wait()
gotit = set()
with open(the_name + '.blast') as b:
for line in b:
query, subject, ident, length = line.split()[:4]
ident = float(ident)
length = int(length)
if ident >= min_ident and length >= min_length * len_dict[
subject]:
gotit.add(subject)
for num, i in enumerate(gene_list):
if i in gotit:
colour = "#5ba965"
else:
colour = "#cb5b4c"
n.add_face(RectFace(20, 20, colour, colour),
column=num + starting_col,
position="aligned")
# for num, i in enumerate(gene_list):
# x = (starting_col + num) * 200
# svg.writeString(i, x+50, 20, 12)
# y = 30
# svg.drawOutRect(x + 50, y, 12, 12, strtorgb('#5ba965'), strtorgb('#5ba965'), lt=0)
# svg.writeString('present', x + 70, y + 12, 12)
# y = 50
# svg.drawOutRect(x + 50, y, 12, 12, strtorgb('#cb5b4c'), strtorgb('#cb5b4c'), lt=0)
# svg.writeString('absent', x + 70, y + 12, 12)
# Set these to False if you don't want bootstrap/distance values
ts.show_branch_length = label
ts.show_branch_support = bootstrap
ts.show_leaf_name = False
for node in t.traverse():
if node.is_leaf():
node.add_face(AttrFace("name",
fsize=font_size,
ftype=font_type,
tight_text=True,
fgcolor='black'),
column=0,
position="aligned")
ts.margin_left = 20
ts.margin_right = 100
ts.margin_top = 20
ts.margin_bottom = 20
if extend:
ts.draw_guiding_lines = True
ts.scale = the_scale
if not circular is None:
ts.mode = "c"
ts.arc_start = 0
ts.arc_span = 360
if out_file is None:
t.show(tree_style=ts)
else:
t.render(out_file, w=210, units='mm', tree_style=ts)
def heatmap_view(tree, orthologous_groups, save_dir):
"""Generates a heatmap of regulation states in all species."""
light_tree = copy.deepcopy(tree) # Tree copy for the light heatmap
# Heat map settings
rect_face_fgcolor = 'black'
locus_tag_len = max(len(gene.locus_tag) + 5
for ortho_grp in orthologous_groups
for gene in ortho_grp.genes)
rect_face_width = locus_tag_len * 8
light_rect_face_width = 20
rect_face_height = 20
rotation = 90
# Sort orthologous groups by the number of regulated genes in each group
orthologous_groups = filter_and_sort_orthologous_grps(orthologous_groups)
# For each species and its gene in each orthologous group, draw a rectangle
for node, light_node in zip(tree.get_leaves(), light_tree.get_leaves()):
for i, orthologous_grp in enumerate(orthologous_groups, start=1):
#get all orthologs in group
matching_genes = [g for g in orthologous_grp.genes \
if g.genome.strain_name == node.name]
#if there is ortholog
if len(matching_genes) > 0:
# Get the first ortholog from the genome in the group
#this is the one with higher probability of regulation.
#so this probability will be displayed for the group
gene = matching_genes[0]
p_regulation = gene.operon.regulation_probability
p_notregulation = 1.0 - p_regulation
p_absence = 0
# No ortholog from this genome
else:
gene = None
p_regulation = 0
p_notregulation = 0
p_absence = 1
# Color of the rectangle is based on probabilities
rect_face_bgcolor = rgb2hex(
p_notregulation, p_regulation, p_absence)
rect_face_text = ('%s [%d]' % (gene.locus_tag, gene.operon.operon_id)
if gene else '')
rect_face_label = {'text': rect_face_text,
'font': 'Courier',
'fontsize': 8,
'color': 'black'}
# Create the rectangle
rect_face = RectFace(rect_face_width, rect_face_height,
rect_face_fgcolor, rect_face_bgcolor,
label=rect_face_label)
light_rect_face = RectFace(light_rect_face_width, rect_face_height,
rect_face_fgcolor, rect_face_bgcolor,
label='')
rect_face.rotation = -rotation
light_rect_face.rotation = -rotation
# Add the rectangle to the corresponding column
node.add_face(rect_face, column=i, position='aligned')
light_node.add_face(light_rect_face, column=i, position='aligned')
ts = TreeStyle()
# Add orthologous group descriptions
descriptions = ['-'.join([grp.description, str(grp.NOGs)]) for grp in orthologous_groups]
max_description_len = max(map(len, descriptions))
descriptions = [
'[%d]' % i + description + ' '*(max_description_len-len(description))
for i, description in enumerate(descriptions, start=1)]
for i, description in enumerate(descriptions, start=1):
text_face = TextFace(description, ftype='Courier')
text_face.hz_align = 1
text_face.vt_align = 1
text_face.rotation = -rotation
ts.aligned_header.add_face(text_face, column=i)
# Rotate the generated heatmap.
ts.margin_left = 10
ts.margin_top = 20
ts.rotation = rotation
ts.show_scale = False
# For some reason, it can't render to PDF in color
tree.render(os.path.join(save_dir, 'heatmap.svg'), tree_style=ts)
light_tree.render(os.path.join(save_dir, 'heatmap_light.svg'), tree_style=ts)
def my_layout(node):
circle_color = 'lightgray' if colormap is None or node.name not in colormap else colormap[
node.name]
text_color = 'black'
if isinstance(circle_color, str):
if isolabel and hasattr(node, 'isotype'):
nl = ''.join(
sorted(set([ISO_SHORT[iss] for iss in node.isotype]),
key=lambda x: ISO_TYPE_charORDER[x]))
else:
nl = str(node.frequency)
C = CircleFace(radius=max(3, 10 * scipy.sqrt(node.frequency)),
color=circle_color,
label={
'text': nl,
'color': text_color
} if node.frequency > 0 else None)
C.rotation = -90
C.hz_align = 1
faces.add_face_to_node(C, node, 0)
else:
P = PieChartFace(
[100 * x / node.frequency for x in circle_color.values()],
2 * 10 * scipy.sqrt(node.frequency),
2 * 10 * scipy.sqrt(node.frequency),
colors=[(color if color != 'None' else 'lightgray')
for color in list(circle_color.keys())],
line_color=None)
T = TextFace(' '.join(
[str(x) for x in list(circle_color.values())]),
tight_text=True)
T.hz_align = 1
T.rotation = -90
faces.add_face_to_node(P, node, 0, position='branch-right')
faces.add_face_to_node(T, node, 1, position='branch-right')
if idlabel:
T = TextFace(node.name, tight_text=True, fsize=6)
T.rotation = -90
T.hz_align = 1
faces.add_face_to_node(
T,
node,
1 if isinstance(circle_color, str) else 2,
position='branch-right')
elif isolabel and hasattr(node, 'isotype') and False:
iso_name = ''.join(
sorted(set([ISO_SHORT[iss] for iss in node.isotype]),
key=lambda x: ISO_TYPE_charORDER[x]))
#T = TextFace(iso_name, tight_text=True, fsize=6)
#T.rotation = -90
#T.hz_align = 1
#faces.add_face_to_node(T, node, 1 if isinstance(circle_color, str) else 2, position='branch-right')
C = CircleFace(radius=max(3, 10 * scipy.sqrt(node.frequency)),
color=circle_color,
label={
'text': iso_name,
'color': text_color
} if node.frequency > 0 else None)
C.rotation = -90
C.hz_align = 1
faces.add_face_to_node(C, node, 0)
def plot_tree_barplot(tree_file, taxon2mlst, header_list):
'''
display one or more barplot
:param tree_file:
:param taxon2value_list:
:param exclude_outgroup:
:param bw_scale:
:param barplot2percentage: list of bool to indicates if the number are percentages and the range should be set to 0-100
:return:
'''
import matplotlib.cm as cm
from matplotlib.colors import rgb2hex
import matplotlib as mpl
mlst_list = list(set(taxon2mlst.values()))
mlst2color = dict(zip(mlst_list, get_spaced_colors(len(mlst_list))))
mlst2color['-'] = 'white'
if isinstance(tree_file, Tree):
t1 = tree_file
else:
t1 = Tree(tree_file)
# Calculate the midpoint node
R = t1.get_midpoint_outgroup()
# and set it as tree outgroup
t1.set_outgroup(R)
tss = TreeStyle()
value = 1
tss.draw_guiding_lines = True
tss.guiding_lines_color = "gray"
tss.show_leaf_name = False
cmap = cm.YlGnBu #YlOrRd#OrRd
scale_list = []
max_value_list = []
for i, lf in enumerate(t1.iter_leaves()):
#if taxon2description[lf.name] == 'Pirellula staleyi DSM 6068':
# lf.name = 'Pirellula staleyi DSM 6068'
# continue
if i == 0:
# header
col_add = 0
#lf.add_face(n, column, position="aligned")
n = TextFace('MLST')
n.margin_top = 1
n.margin_right = 2
n.margin_left = 2
n.margin_bottom = 1
n.rotation = 90
n.inner_background.color = "white"
n.opacity = 1.
n.hz_align = 2
n.vt_align = 2
tss.aligned_header.add_face(n, col_add + 1)
try:
#if lf.name in leaf2mlst or int(lf.name) in leaf2mlst:
n = TextFace(' %s ' % taxon2mlst[int(lf.name)])
n.inner_background.color = 'white'
m = TextFace(' ')
m.inner_background.color = mlst2color[taxon2mlst[int(lf.name)]]
except:
n = TextFace(' na ')
n.inner_background.color = "grey"
m = TextFace(' ')
m.inner_background.color = "white"
n.opacity = 1.
n.margin_top = 2
n.margin_right = 2
n.margin_left = 0
n.margin_bottom = 2
m.margin_top = 2
m.margin_right = 0
m.margin_left = 2
m.margin_bottom = 2
lf.add_face(m, 0, position="aligned")
lf.add_face(n, 1, position="aligned")
n = TextFace(lf.name, fgcolor="black", fsize=12, fstyle='italic')
lf.add_face(n, 0)
for n in t1.traverse():
nstyle = NodeStyle()
if n.support < 1:
nstyle["fgcolor"] = "black"
nstyle["size"] = 6
n.set_style(nstyle)
else:
nstyle["fgcolor"] = "red"
nstyle["size"] = 0
n.set_style(nstyle)
return t1, tss
def plot_tree_barplot(tree_file,
taxon2value_list_barplot,
header_list,
taxon2set2value_heatmap=False,
header_list2=False,
column_scale=True,
general_max=False,
barplot2percentage=False,
taxon2mlst=False):
'''
display one or more barplot
:param tree_file:
:param taxon2value_list:
:param exclude_outgroup:
:param bw_scale:
:param barplot2percentage: list of bool to indicates if the number are percentages and the range should be set to 0-100
:return:
'''
import matplotlib.cm as cm
from matplotlib.colors import rgb2hex
import matplotlib as mpl
if taxon2mlst:
mlst_list = list(set(taxon2mlst.values()))
mlst2color = dict(zip(mlst_list, get_spaced_colors(len(mlst_list))))
mlst2color['-'] = 'white'
if isinstance(tree_file, Tree):
t1 = tree_file
else:
t1 = Tree(tree_file)
# Calculate the midpoint node
R = t1.get_midpoint_outgroup()
# and set it as tree outgroup
t1.set_outgroup(R)
tss = TreeStyle()
value = 1
tss.draw_guiding_lines = True
tss.guiding_lines_color = "gray"
tss.show_leaf_name = False
if column_scale and header_list2:
import matplotlib.cm as cm
from matplotlib.colors import rgb2hex
import matplotlib as mpl
column2scale = {}
for column in header_list2:
values = taxon2set2value_heatmap[column].values()
norm = mpl.colors.Normalize(vmin=min(values), vmax=max(values))
cmap = cm.OrRd
m = cm.ScalarMappable(norm=norm, cmap=cmap)
column2scale[column] = m
cmap = cm.YlGnBu #YlOrRd#OrRd
values_lists = taxon2value_list_barplot.values()
scale_list = []
max_value_list = []
for n, header in enumerate(header_list):
#print 'scale', n, header
data = [float(i[n]) for i in values_lists]
if barplot2percentage is False:
max_value = max(data) #3424182#
min_value = min(data) #48.23
else:
if barplot2percentage[n] is True:
max_value = 100
min_value = 0
else:
max_value = max(data) #3424182#
min_value = min(data) #48.23
norm = mpl.colors.Normalize(vmin=min_value, vmax=max_value)
m1 = cm.ScalarMappable(norm=norm, cmap=cmap)
scale_list.append(m1)
if not general_max:
max_value_list.append(float(max_value))
else:
max_value_list.append(general_max)
for i, lf in enumerate(t1.iter_leaves()):
#if taxon2description[lf.name] == 'Pirellula staleyi DSM 6068':
# lf.name = 'Pirellula staleyi DSM 6068'
# continue
if i == 0:
col_add = 0
if taxon2mlst:
header_list = ['MLST'] + header_list
for col, header in enumerate(header_list):
#lf.add_face(n, column, position="aligned")
n = TextFace(' ')
n.margin_top = 1
n.margin_right = 2
n.margin_left = 2
n.margin_bottom = 1
n.rotation = 90
n.inner_background.color = "white"
n.opacity = 1.
n.hz_align = 2
n.vt_align = 2
tss.aligned_header.add_face(n, col_add + 1)
n = TextFace('%s' % header)
n.margin_top = 1
n.margin_right = 2
n.margin_left = 2
n.margin_bottom = 2
n.rotation = 270
n.inner_background.color = "white"
n.opacity = 1.
n.hz_align = 2
n.vt_align = 1
tss.aligned_header.add_face(n, col_add)
col_add += 2
if header_list2:
for col, header in enumerate(header_list2):
n = TextFace('%s' % header)
n.margin_top = 1
n.margin_right = 20
n.margin_left = 2
n.margin_bottom = 1
n.rotation = 270
n.hz_align = 2
n.vt_align = 2
n.inner_background.color = "white"
n.opacity = 1.
tss.aligned_header.add_face(n, col + col_add)
if taxon2mlst:
try:
#if lf.name in leaf2mlst or int(lf.name) in leaf2mlst:
n = TextFace(' %s ' % taxon2mlst[int(lf.name)])
n.inner_background.color = 'white'
m = TextFace(' ')
m.inner_background.color = mlst2color[taxon2mlst[int(lf.name)]]
except:
n = TextFace(' na ')
n.inner_background.color = "grey"
m = TextFace(' ')
m.inner_background.color = "white"
n.opacity = 1.
n.margin_top = 2
n.margin_right = 2
n.margin_left = 0
n.margin_bottom = 2
m.margin_top = 2
m.margin_right = 0
m.margin_left = 2
m.margin_bottom = 2
lf.add_face(m, 0, position="aligned")
lf.add_face(n, 1, position="aligned")
col_add = 2
else:
col_add = 0
try:
val_list = taxon2value_list_barplot[lf.name]
except:
if not taxon2mlst:
val_list = ['na'] * len(header_list)
else:
val_list = ['na'] * (len(header_list) - 1)
for col, value in enumerate(val_list):
# show value itself
try:
n = TextFace(' %s ' % str(value))
except:
n = TextFace(' %s ' % str(value))
n.margin_top = 1
n.margin_right = 5
n.margin_left = 10
n.margin_bottom = 1
n.inner_background.color = "white"
n.opacity = 1.
lf.add_face(n, col_add, position="aligned")
# show bar
try:
color = rgb2hex(scale_list[col].to_rgba(float(value)))
except:
color = 'white'
try:
percentage = (value / max_value_list[col]) * 100
#percentage = value
except:
percentage = 0
try:
maximum_bar = (
(max_value_list[col] - value) / max_value_list[col]) * 100
except:
maximum_bar = 0
#maximum_bar = 100-percentage
b = StackedBarFace([percentage, maximum_bar],
width=100,
height=10,
colors=[color, "white"])
b.rotation = 0
b.inner_border.color = "grey"
b.inner_border.width = 0
b.margin_right = 15
b.margin_left = 0
lf.add_face(b, col_add + 1, position="aligned")
col_add += 2
if taxon2set2value_heatmap:
shift = col + col_add + 1
i = 0
for col, col_name in enumerate(header_list2):
try:
value = taxon2set2value_heatmap[col_name][lf.name]
except:
try:
value = taxon2set2value_heatmap[col_name][int(lf.name)]
except:
value = 0
if int(value) > 0:
if int(value) > 9:
n = TextFace(' %i ' % int(value))
else:
n = TextFace(' %i ' % int(value))
n.margin_top = 1
n.margin_right = 1
n.margin_left = 20
n.margin_bottom = 1
n.fgcolor = "white"
n.inner_background.color = rgb2hex(
column2scale[col_name].to_rgba(
float(value))) #"orange"
n.opacity = 1.
lf.add_face(n, col + col_add, position="aligned")
i += 1
else:
n = TextFace(' ') #% str(value))
n.margin_top = 1
n.margin_right = 1
n.margin_left = 20
n.margin_bottom = 1
n.inner_background.color = "white"
n.opacity = 1.
lf.add_face(n, col + col_add, position="aligned")
n = TextFace(lf.name, fgcolor="black", fsize=12, fstyle='italic')
lf.add_face(n, 0)
for n in t1.traverse():
nstyle = NodeStyle()
if n.support < 1:
nstyle["fgcolor"] = "black"
nstyle["size"] = 6
n.set_style(nstyle)
else:
nstyle["fgcolor"] = "red"
nstyle["size"] = 0
n.set_style(nstyle)
return t1, tss
def make_cluster_tree(tree_file: str,
matrix: str,
out_file: str,
outgroup: Optional[List[str]] = None) -> None:
"""Draw a tree with cluster absence/presence information from an existing
tree file and absence/presence matrix, and save it as an image under the
supplied file name.
Arguments:
tree_file: the name of the file containing the tree to annotate
matrix: a comma- or tab-separated absence/presence matrix
out_file: the name under which to save the resulting image
outgroup: the organism(s) to use as an outgroup, if any
"""
# ClusterTree needs tab-separated, but that can't be exported cleanly
matrix = matrix.replace(",", "\t")
# tree with clustering analysis
tree = ClusterTree(tree_file, text_array=matrix)
# rerooting the tree
if outgroup:
ancestor = tree.get_common_ancestor(outgroup)
tree.set_outgroup(ancestor)
tree.ladderize(direction=1)
# set drawing line width to 2
my_node_style = NodeStyle()
my_node_style["vt_line_width"] = 2
my_node_style["hz_line_width"] = 2
my_node_style["size"] = 5
# layout function
def sel_mylayout(node):
node.set_style(my_node_style)
if node.is_leaf():
# add names in larger font + italics
species_name = AttrFace("name", fsize=12, fstyle="italic")
add_face_to_node(species_name,
node,
column=0,
position="branch-right")
# add absence/presence matrix
for i, value in enumerate(getattr(node, "profile", [])):
if value > 0:
color = "#FF0000"
else:
color = "#EEEEEE"
my_face = CircleFace(8, color, style="circle")
my_face.margin_right = 3
my_face.margin_bottom = 3
add_face_to_node(my_face, node, position="aligned", column=i)
# Use my layout to visualize the tree
my_tree_style = TreeStyle()
# Add header
for j, name in enumerate(tree.arraytable.colNames):
name_face = TextFace(name, fsize=11)
name_face.rotation = -90
name_face.hz_align = 1
name_face.vt_align = 1
name_face.margin_bottom = 10
my_tree_style.aligned_header.add_face(name_face, column=j)
my_tree_style.scale_length = 0.1
# myTreeStyle.show_branch_support = True
# don't auto-show leaf names, since we dealt with that above
my_tree_style.show_leaf_name = False
# set layout function for my_tree_style
my_tree_style.layout_fn = sel_mylayout
#tree.render(out_file, w=183, units="mm", dpi=600, tree_style=my_tree_style)
tree.render(out_file, dpi=600, tree_style=my_tree_style)
