def custom_layout(node):
if node.is_leaf():
aligned_name_face = TextFace(node.name, fgcolor='olive', fsize=14)
add_face_to_node(aligned_name_face, node, column=2, position='aligned')
name_face = TextFace(node.name, fgcolor='#333333', fsize=11)
add_face_to_node(name_face, node, column=2, position='branch-right')
node.img_style['size'] = 0
if node.name in tip2info:
# For some reason img urls are very slow!
#img_face = ImgFace(tip2info[node.name][0], is_url=True)
#add_face_to_node(img_face, node, column=4, position='branch-right')
habitat_face = TextFace(tip2info[node.name][2], fsize=11, fgcolor='white')
habitat_face.background.color = 'steelblue'
habitat_face.margin_left = 3
habitat_face.margin_top = 3
habitat_face.margin_right = 3
habitat_face.margin_bottom = 3
add_face_to_node(habitat_face, node, column=3, position='aligned')
else:
node.img_style['size'] = 4
node.img_style['shape'] = 'square'
if node.name:
name_face = TextFace(node.name, fgcolor='grey', fsize=10)
name_face.margin_bottom = 2
add_face_to_node(name_face, node, column=0, position='branch-top')
if node.support:
support_face = TextFace(node.support, fgcolor='indianred', fsize=10)
add_face_to_node(support_face, node, column=0, position='branch-bottom')
def custom_layout(node):
if node.is_leaf():
aligned_name_face = TextFace(node.name, fgcolor='olive', fsize=14)
add_face_to_node(aligned_name_face, node, column=2, position='aligned')
name_face = TextFace(node.name, fgcolor='#333333', fsize=11)
add_face_to_node(name_face, node, column=2, position='branch-right')
node.img_style['size'] = 0
if (node.name in tip2info) and (node.name in image_checker):
# image
img_face = ImgFace(tip2info[node.name][0], is_url=True)
add_face_to_node(img_face, node, column=4, position='branch-right')
habitat_face = TextFace(tip2info[node.name][2], fsize=11, fgcolor='white')
habitat_face.background.color = 'steelblue'
habitat_face.margin_left = 3
habitat_face.margin_top = 3
habitat_face.margin_right = 3
habitat_face.margin_bottom = 3
add_face_to_node(habitat_face, node, column=3, position='aligned')
else:
node.img_style['size'] = 4
node.img_style['shape'] = 'square'
if node.name:
name_face = TextFace(node.name, fgcolor='grey', fsize=10)
name_face.margin_bottom = 2
add_face_to_node(name_face, node, column=0, position='branch-top')
if node.support:
support_face = TextFace(node.support, fgcolor='indianred', fsize=10)
add_face_to_node(support_face, node, column=0, position='branch-bottom')
def styleFace(val):
x = TextFace(val)
x.margin_bottom = 5
x.margin_right = 10
x.rotation = 270
x.fsize = 6
return x
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 traitTree(traits, mapper, outDir):
### Take dict of traits and [R,G,B]-returning function
### Draw a tree with the continuous trait painted on via a colormapping function
def rgb2hex(r, g, b):
hex = "#{:02x}{:02x}{:02x}".format(r, g, b)
return hex
for n in tree.traverse():
if n.is_leaf():
n.set_style(nstyle_L)
n.add_face(TextFace(str(n.name)), column=0, position="aligned")
else:
n.set_style(nstyle)
#nd = TextFace(str(n.ND)) # label with node ID
nd = TextFace(str(int(
traits[n.ND]))) # label with rounded continuous trait value
nd.background.color = rgb2hex(*[
int(val) for val in mapper(traits[n.ND], gamma=0.8, scaleMax=255)
]) # setup for wl2RGB
nd.margin_right = 2
nd.margin_top = 1
nd.margin_left = 2
nd.margin_bottom = 1
nd.border.width = 1
n.add_face(nd, column=0, position="float")
n.add_face(TextFace(" "), column=0, position="branch-bottom")
#outFile = args.output + "/test_trees/cont_trait.pdf"
outFile = outDir + "/cont_trait.pdf"
tree.render(outFile, tree_style=tree_style)
print >> sys.stderr, outFile
def draw_tree(tree):
tree_copy = tree.copy("newick-extended")
tree_copy.add_feature("i", tree.i)
tree_copy.add_feature("Condition", tree.Condition)
for n in tree_copy.traverse():
if n.is_leaf():
n.set_style(nstyle)
n.add_face(TextFace(str(n.name)), column=0, position="aligned")
else:
n.set_style(nstyle)
nd = TextFace(str(n.i))
nd.background.color = condi_color_dic[str(n.Condition)]
nd.margin_right = 2
nd.margin_top = 1
nd.margin_left = 2
nd.margin_bottom = 1
nd.border.width = 1
if add_transition:
if hasattr(n, "Transition"):
nd.border.color = "red"
nd.border.width = 2
n.add_face(nd, column=0, position="float")
n.add_face(TextFace(" "), column=0, position="branch-bottom")
return tree_copy
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 testTrees(scenarios, outDir):
### Draw test trees. This is a modified version of the test routine in pcoc_num_tree.py, stuffed in a for loop
for cutoff in sorted(scenarios.keys()):
tree = init_tree(args.tree)
# not mucking with additive trees yet; ultrametrize the tree and normalize to length 1
tree.convert_to_ultrametric(tree_length=1)
manual_mode_nodes = {}
manual_mode_nodes = {"T": [], "C": []}
p_events = scenarios[cutoff].strip().split("/")
for e in p_events:
l_e = map(int, e.split(","))
manual_mode_nodes["T"].append(l_e[0])
manual_mode_nodes["C"].extend(l_e[1:])
for n in tree.traverse():
if n.is_leaf():
n.set_style(nstyle_L)
n.add_face(TextFace(str(n.name)), column=0, position="aligned")
else:
n.set_style(nstyle)
nd = TextFace(str(n.ND))
if manual_mode_nodes:
if n.ND in manual_mode_nodes["T"]:
nd.background.color = "red"
elif n.ND in manual_mode_nodes["C"]:
nd.background.color = "orange"
else:
nd.background.color = "white"
else:
nd.background.color = "white"
nd.background.color = "white"
nd.margin_right = 2
nd.margin_top = 1
nd.margin_left = 2
nd.margin_bottom = 1
nd.border.width = 1
n.add_face(nd, column=0, position="float")
n.add_face(TextFace(" "), column=0, position="branch-bottom")
# if --float set, limit number of digits in filename
if args.float:
outFile = str(cutoff).replace(
'.', '_')[:np.min([args.float, len(str(cutoff))])] + ".pdf"
else:
outFile = str(cutoff).replace('.', '_') + ".pdf"
# prepend path to filename
outFile = outDir + '/' + outFile
tree.render(outFile, tree_style=tree_style)
print >> sys.stderr, outFile
def add_t(node):
nd = TextFace("-")
nd.fsize = 4
nd.background.color = "black"
nd.margin_right = 0
nd.margin_top = 0
nd.margin_left = 0
nd.margin_bottom = 0
nd.border.width = 1
nd2 = TextFace(" ")
nd2.fsize = 4
node.add_face(nd, column=0, position = "float")
node.add_face(nd2, column=1, position = "float")
def custom_layout(node):
if node.is_leaf():
#FACES IN LEAFS
#Add face with node.name
total_name = (node.name)
aligned_name_face = TextFace(total_name, fgcolor='gray', fsize=11)
add_face_to_node(aligned_name_face,
node,
column=0,
position='branch-right')
#if tree has an alignment, add a face with alignment
if node.sequence:
seqFace = SeqMotifFace(node.sequence, gap_format="blank")
add_face_to_node(seqFace, node, column=1, position="aligned")
else:
#FACES IN INTERNAL NODES
#Draws nodes as small blue squares of diameter equal to 3 pixels
node.img_style['size'] = 3
node.img_style['shape'] = 'square'
#if internal node has a name, add a face
if node.name:
name_face = TextFace(node.name, fgcolor='brown', fsize=10)
name_face.margin_bottom = 1
add_face_to_node(name_face, node, column=0, position='branch-top')
#if internal node has support (ej. boostrap), add a face
if node.support:
support_face = TextFace(node.support, fgcolor='indianred', fsize=8)
support_face.margin_bottom = 1
add_face_to_node(support_face,
node,
column=0,
position='branch-bottom')
def add_continuous_colorscale_legend(self,
title,
min_val,
max_val,
scale):
self.tss.legend.add_face(TextFace(f"{title}", fsize = 4 * self.text_scale), column=0)
if min_val != max_val:
n = TextFace(" " * int(self.text_scale), fsize = 4 * self.text_scale)
n.margin_top = 1
n.margin_right = 1
n.margin_left = 10
n.margin_bottom = 1
n.inner_background.color = rgb2hex(scale[0].to_rgba(float(max_val)))
n2 = TextFace(" " * int(self.text_scale), fsize = 4 * self.text_scale)
n2.margin_top = 1
n2.margin_right = 1
n2.margin_left = 10
n2.margin_bottom = 1
n2.inner_background.color = rgb2hex(scale[0].to_rgba(float(min_val)))
self.tss.legend.add_face(n, column=1)
self.tss.legend.add_face(TextFace(f"{max_val} % (max)", fsize = 4 * self.text_scale), column=2)
self.tss.legend.add_face(n2, column=1)
self.tss.legend.add_face(TextFace(f"{min_val} % (min)", fsize = 4 * self.text_scale), column=2)
else:
n2 = TextFace(" " * int(self.text_scale), fsize = 4 * self.text_scale)
n2.margin_top = 1
n2.margin_right = 1
n2.margin_left = 10
n2.margin_bottom = 1
n2.inner_background.color = rgb2hex(scale[0].to_rgba(float(min_val)))
self.tss.legend.add_face(n2, column=0)
self.tss.legend.add_face(TextFace(f"{max_val} % Id", fsize = 4 * self.text_scale), column=1)
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 custom_layout(node):
if node.is_leaf():
aligned_name_face = TextFace(node.name, fgcolor='olive', fsize=14)
add_face_to_node(aligned_name_face, node, column=2, position='aligned')
name_face = TextFace(node.name, fgcolor='#333333', fsize=11)
add_face_to_node(name_face, node, column=2, position='branch-right')
node.img_style['size'] = 0
else:
node.img_style['size'] = 4
node.img_style['shape'] = 'square'
if node.name:
name_face = TextFace(node.name, fgcolor='grey', fsize=10)
name_face.margin_bottom = 2
add_face_to_node(name_face, node, column=0, position='branch-top')
if node.support:
support_face = TextFace(node.support, fgcolor='indianred', fsize=10)
add_face_to_node(support_face, node, column=0, position='branch-bottom')
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 testTrees(scenarios, tree_style):
### Draw test trees. This is a modified version of the test routine in pcoc_num_tree.py, stuffed in a for loop
for cutoff in sorted(scenarios.keys()):
# this keeps attributes from stacking up in the same tree
tree = init_tree(args.tree)
manual_mode_nodes = {}
manual_mode_nodes = {"T": [], "C": []}
p_events = scenarios[cutoff].strip().split("/")
for e in p_events:
l_e = map(int, e.split(","))
manual_mode_nodes["T"].append(l_e[0])
manual_mode_nodes["C"].extend(l_e[1:])
for n in tree.traverse():
if n.is_leaf():
n.set_style(nstyle_L)
n.add_face(TextFace(str(n.name)), column=0, position="aligned")
else:
n.set_style(nstyle)
nd = TextFace(str(n.ND))
if manual_mode_nodes:
if n.ND in manual_mode_nodes["T"]:
nd.background.color = "red"
elif n.ND in manual_mode_nodes["C"]:
nd.background.color = "orange"
else:
nd.background.color = "white"
else:
nd.background.color = "white"
nd.background.color = "white"
nd.margin_right = 2
nd.margin_top = 1
nd.margin_left = 2
nd.margin_bottom = 1
nd.border.width = 1
n.add_face(nd, column=0, position="float")
n.add_face(TextFace(" "), column=0, position="branch-bottom")
outfile = args.output + "/test_trees/" + str(cutoff).replace('.','_') + ".pdf"
tree.render(outfile, tree_style=tree_style)
print >> sys.stderr, outfile
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_categorical_colorscale_legend(self,
title,
scale):
self.tss.legend.add_face(TextFace(f"{title}", fsize = 4 * self.text_scale), column=0)
col = 1
for n,value in enumerate(scale):
n2 = TextFace(" " * int(self.text_scale), fsize = 4 * self.text_scale)
n2.margin_top = 1
n2.margin_right = 1
n2.margin_left = 10
n2.margin_bottom = 1
n2.inner_background.color = scale[value]
self.tss.legend.add_face(n2, column=col)
self.tss.legend.add_face(TextFace(f"{value}", fsize = 4 * self.text_scale), column=col+1)
col+=2
if col>16:
self.tss.legend.add_face(TextFace(f" ", fsize = 4 * self.text_scale), column=0)
col = 1
def style_and_render_tree(self, file_types=["svg"]):
from ete3 import TreeStyle, TextFace, CircleFace
ts = TreeStyle()
title_face = TextFace(f"{genus} {snakemake.config['species']}", fsize=20)
title_face.margin_bottom = 10
ts.title.add_face(title_face, column=0)
ts.branch_vertical_margin = 10
ts.show_leaf_name = True
# Legend
ts.legend.add_face(TextFace(""), column=1)
for category in ["Allowed", "Deviations", "Filtered", "Color"]:
category = TextFace(category, fsize=8, bold=True)
category.margin_bottom = 2
category.margin_right = 40
ts.legend.add_face(category, column=1)
for i, criteria in enumerate(CRITERIA, 2):
title = criteria.replace("_", " ").title()
title = TextFace(title, fsize=8, bold=True)
title.margin_bottom = 2
title.margin_right = 40
cf = CircleFace(4, COLORS[criteria], style="sphere")
cf.margin_bottom = 5
filtered_count = len(
list(filter(None, self.failed_report.criteria == criteria))
)
filtered = TextFace(filtered_count, fsize=8)
filtered.margin_bottom = 5
allowed = TextFace(self.allowed[criteria], fsize=8)
allowed.margin_bottom = 5
allowed.margin_right = 25
# TODO Prevent tolerance from rendering as a float
tolerance = TextFace(self.tolerance[criteria], fsize=8)
tolerance.margin_bottom = 5
ts.legend.add_face(title, column=i)
ts.legend.add_face(allowed, column=i)
ts.legend.add_face(tolerance, column=i)
ts.legend.add_face(filtered, column=i)
ts.legend.add_face(cf, column=i)
for f in file_types:
out_tree = os.path.join(self.paths.qc, "tree.{}".format(f))
self.tree.render(out_tree, tree_style=ts)
def style_and_render_tree(self, file_types=["svg"]):
from ete3 import TreeStyle, TextFace, CircleFace
ts = TreeStyle()
title_face = TextFace(self.name.replace('_', ' '), fsize=20)
title_face.margin_bottom = 10
ts.title.add_face(title_face, column=0)
ts.branch_vertical_margin = 10
ts.show_leaf_name = False
# Legend
ts.legend.add_face(TextFace(""), column=1)
for category in ["Allowed", "Tolerance", "Filtered", "Color"]:
category = TextFace(category, fsize=8, bold=True)
category.margin_bottom = 2
category.margin_right = 40
ts.legend.add_face(category, column=1)
for i, criteria in enumerate(self.criteria, 2):
title = criteria.replace("_", " ").title()
title = TextFace(title, fsize=8, bold=True)
title.margin_bottom = 2
title.margin_right = 40
cf = CircleFace(4, self.colors[criteria], style="sphere")
cf.margin_bottom = 5
filtered_count = len(
list(filter(None, self.failed_report.criteria == criteria)))
filtered = TextFace(filtered_count, fsize=8)
filtered.margin_bottom = 5
allowed = TextFace(self.allowed[criteria], fsize=8)
allowed.margin_bottom = 5
allowed.margin_right = 25
tolerance = TextFace(self.tolerance[criteria], fsize=8)
tolerance.margin_bottom = 5
ts.legend.add_face(title, column=i)
ts.legend.add_face(allowed, column=i)
ts.legend.add_face(tolerance, column=i)
ts.legend.add_face(filtered, column=i)
ts.legend.add_face(cf, column=i)
for f in file_types:
out_tree = os.path.join(self.qc_results_dir, 'tree.{}'.format(f))
self.tree.render(out_tree, tree_style=ts)
self.log.info("tree.{} generated".format(f))
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 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)
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 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
ts = TreeStyle()
ts.show_leaf_name = False
ts.show_branch_length = False
ts.show_branch_support = True
ts.scale = 2
#ts.min_leaf_separation = 3
ts.branch_vertical_margin = 12
ts.legend_position = 4
#ts.legend.add_face(CircleFace(3, "red"), column=0)
mark = TextFace("Outbreak", fsize=10, fgcolor=outbreak_color)
mark.margin_top = 10
mark.margin_right = 10
mark.margin_left = 5
mark.margin_bottom = 10
#ts.legend.add_face(mark, column=1)
mark2 = TextFace("X", fsize=10, fgcolor="black")
# Set some attributes
mark2.margin_top = 0
mark2.margin_right = 1
mark2.margin_left = 1
mark2.margin_bottom = 0
mark2.opacity = 1 # from 0 to 1
mark2.border.width = 1
mark2.background.color = "#F5F5DC"
ts.legend.add_face(mark2, column=0)
mark3 = TextFace("Selected branches", fsize=10, fgcolor="black")
mark3.margin_top = 2
manual_mode_nodes["T"].append(l_e[0])
manual_mode_nodes["C"].extend(l_e[1:])
for n in tree.traverse():
if n.is_leaf():
n.set_style(nstyle_L)
n.add_face(TextFace(str(n.name)), column=0, position="aligned")
else:
n.set_style(nstyle)
nd = TextFace(str(n.ND))
if manual_mode_nodes:
if n.ND in manual_mode_nodes["T"]:
nd.background.color = "red"
elif n.ND in manual_mode_nodes["C"]:
nd.background.color = "orange"
else:
nd.background.color = "white"
else:
nd.background.color = "white"
nd.margin_right = 2
nd.margin_top = 1
nd.margin_left = 2
nd.margin_bottom = 1
nd.border.width = 1
n.add_face(nd, column=0, position="float")
n.add_face(TextFace(" "), column=0, position="branch-bottom")
tree.render(args.output, tree_style=tree_style)
print args.output
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 custom_layout(self,node):
if node.is_leaf():
aligned_name_face = TextFace(node.name, fgcolor='olive', fsize=12)
aligned_name_face.margin_top = 5
aligned_name_face.margin_right = 5
aligned_name_face.margin_left = 5
aligned_name_face.margin_bottom = 5
aligned_name_face.hz_align = 0 #0 = left, 1 = center, 2 = right
add_face_to_node(aligned_name_face, node, column=2, position='aligned')
#name_face = TextFace(node.name, fgcolor='#333333', fsize=11)
#name_face.margin_top = 3
#name_face.margin_right = 3
#name_face.margin_left = 3
#name_face.margin_bottom = 3
#add_face_to_node(name_face, node, column=2, position='branch-right')
node.img_style['size'] = 0
#---------------------------------------------
#displaying extra categorical and numeric data
if (node.name in self._tip2info):
column_no = 3
for headerIndex, dataheader in enumerate(self._tip2headers):
extra_data = self._tip2info[node.name][headerIndex]
if isinstance( extra_data, ( int, float ) ):
extra_face = BarChartFace([extra_data], width=100,height=25,colors=[self._tip2color[node.name][headerIndex]],labels=[dataheader],min_value=0.0,max_value=self._tip_max)
else:
extra_face = TextFace(extra_data, fsize=11, fgcolor='black')
extra_face.background.color = self._tip2color[node.name][headerIndex]
extra_face.margin_left = 5
extra_face.margin_top = 5
extra_face.margin_right = 5
extra_face.margin_bottom = 5
add_face_to_node(extra_face, node, column=column_no, position='aligned')
#add_face_to_node(extra_face, node, column=column_no, aligned = True, position='branch-right')
column_no += 1
else:
#print "No data available"
column_no = 3
for headerIndex, dataheader in enumerate(self._tip2headers):
extra_face = TextFace("No data available", fsize=10, fgcolor='black')
extra_face.margin_left = 5
extra_face.margin_top = 5
extra_face.margin_right = 5
extra_face.margin_bottom = 5
add_face_to_node(extra_face, node, column=column_no, position='aligned')
column_no += 1
image_col_no = column_no
#----------------------------------------------
if (node.name in self._img_chk_list):
if self._img_data_dic[node.name] is not None:
img_face = ImgFace(self._img_data_dic[node.name], is_url=True)
#img_face = ImgFace(self._tip2info[node.name][0], is_url=True)
#img_path = os.path.join("file:///home/tayeen/TayeenFolders/TreeViewer/WebTreeApp/newplugin_test/data/", "328653.jpg")
#img_face = ImgFace(img_path, is_url=True)
img_face.margin_top = 10
img_face.margin_right = 10
img_face.margin_left = 10
img_face.margin_bottom = 10
#add_face_to_node(img_face, node, column=3, position='branch-right')
#add_face_to_node(img_face, node, column=3, aligned= True, position='branch-right')
else:
img_path = os.path.join("file://"+image_path, "ina.jpg")
img_face = ImgFace(img_path, is_url=True)
#add_face_to_node(img_face, node, column=5, position='branch-right')
add_face_to_node(img_face, node, column=image_col_no, position='aligned')
else: #node is not a leaf
node.img_style['size'] = 4
node.img_style['shape'] = 'square'
if node.name and self._custom_options["draw_internal"]:
name_face = TextFace(node.name, fgcolor='grey', fsize=10)
name_face.margin_top = 4
name_face.margin_right = 4
name_face.margin_left = 4
name_face.margin_bottom = 4
add_face_to_node(name_face, node, column=0, position='branch-top')
if node.name in self._node2label:
label_face = TextFace(self._node2label[node.name], fgcolor='DarkGreen', fsize=10)
label_face.margin_top = 4
label_face.margin_right = 4
label_face.margin_left = 4
label_face.margin_bottom = 4
add_face_to_node(label_face, node, column=0, position="branch-top")
if node.support and self._custom_options["draw_support"]:
support_face = TextFace(node.support, fgcolor='indianred', fsize=10)
support_face.margin_top = 4
support_face.margin_right = 4
support_face.margin_left = 4
support_face.margin_bottom = 4
add_face_to_node(support_face, node, column=0, position='branch-bottom')
if hasattr(node, "hide") and int(node.hide) == 1:
node.img_style["draw_descendants"]= False
collapsed_face = faces.TextFace(" %s collapsed leaves." %len(node), \
fsize=10, fgcolor="#444", ftype="Arial")
faces.add_face_to_node(collapsed_face, node, 0)
else:
node.img_style["draw_descendants"] = True
# Parse node features features and conver them into styles. This must be done like this, since current ete version
#does not allow modifying style outside the layout function.
if hasattr(node, "bsize"):
node.img_style["size"]= int(node.bsize)
if hasattr(node, "shape"):
node.img_style["shape"]= node.shape
if hasattr(node, "bgcolor"):
node.img_style["bgcolor"]= node.bgcolor
if hasattr(node, "fgcolor"):
node.img_style["fgcolor"]= node.fgcolor
#parse all nodes features
if hasattr(node, "bh_bgcolor"):
node.img_style["bgcolor"]= node.bh_bgcolor
if hasattr(node, "bh_size"):
node.img_style["size"]= node.bh_size
if hasattr(node, "lh_color"):
node.img_style['hz_line_color'] = node.lh_color
node.img_style["vt_line_color"] = node.lh_color
if hasattr(node, "lh_width"):
node.img_style['hz_line_width'] = node.lh_width
node.img_style['vt_line_width'] = node.lh_width
if hasattr(node, "lh_width") and hasattr(node, "lh_color"):
for n in node.iter_descendants():
n.img_style['hz_line_color'] = node.lh_color
n.img_style["vt_line_color"] = node.lh_color
n.img_style['hz_line_width'] = node.lh_width
n.img_style['vt_line_width'] = node.lh_width
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_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 visualize_phylogeny(gene_dict, context_file):
#Read in tree and assign additional information to each leaf
t = Tree(context_file[0].replace('.fna', '.unique.tree'))
for node in t.traverse():
if node.is_leaf():
id = node.name.split('__')[1]
node.add_features(organism=gene_dict[id]['organism'])
node.add_features(assembly=gene_dict[id]['assembly'])
node.add_features(pident=gene_dict[id]['perc_id'])
if args.compressed == True:
node.add_features(cluster_size=gene_dict[id]['cluster_size'])
#Create dictionary to append motifs to
motif_dict = {}
#Create keyword lists to set gene color
tnps = [
'iscr',
'transpos',
'tnp',
'insertion',
'-like',
]
ints = ['inti', 'integrase', 'xerc', 'xerd']
mobiles=['secretion', 'mobiliza', 'moba', 'mobb', 'mobc', 'mobl', 'plasmid', 'relaxase',\
'conjugation', 'type iv']
res = [
'lactam', 'aminoglyco', 'fluoroquinolo', 'tetracyclin', 'macrolid',
'carbapenem'
]
print('decorating the tree...')
#Create motifs for each gene associated with a leaf
for leaf in t.traverse():
if leaf.is_leaf():
#traverse through environment genes for the respective sequence
for key, value in gene_dict.items():
motifs = []
#Assign start and end position for annotated gene
gene_start = gene_dict[key]['start']
gene_end = gene_dict[key]['stop']
#Sort such that the greater number is end and smaller is start
if gene_start > gene_end:
gene_end = gene_dict[key]['start']
gene_start = gene_dict[key]['stop']
#Append motif for annotated gene
gene_motif=[gene_start, gene_end,'()', \
2, 10, 'red', 'red', 'arial|10|black|'+str(gene_dict[key]['name'])]
if not str(gene_dict[key]['frame']).startswith('-'):
ori_motif = [
gene_end, gene_end + 10, '>', 2, 10, 'red', 'red', None
]
else:
ori_motif=[gene_start-10, gene_start, '<', 2, 10, \
'red', 'red', None]
motifs.extend([gene_motif, ori_motif])
for key2, value2 in value['env_genes'].items():
#Set color, default is orange
color = 'orange'
if any(keyword in value2['env_name'].lower()
for keyword in tnps):
color = 'violet'
if any(keyword in value2['env_name'].lower()
for keyword in ints):
color = 'yellow'
if any(keyword in value2['env_name'].lower()
for keyword in mobiles):
color = 'green'
if any(keyword in value2['env_name'].lower()
for keyword in res):
color = 'red'
if 'hypothetical' in value2['env_name']:
color = 'grey'
#Create motif for one env gene at a time and append to motif list
motif=[value2['env_start'], value2['env_stop'], '()', 2, 10, color, color, \
'arial|10|black|'+str(value2['env_name'])]
#Set condition: If env gene != annotated gene, append motif
arg_pos = {
i
for i in range(int(gene_motif[0]), int(gene_motif[1]))
}
env_pos = {i for i in range(int(motif[0]), int(motif[1]))}
#Calculate overlap percentage between annotated gene and env gene
total_overlap = float(len(arg_pos.intersection(env_pos)))
overlap_perc = float(
total_overlap / int(gene_dict[key]['length'])) * 100
if overlap_perc <= 70.0:
motifs.append(motif)
#Create additional motif to show gene orientation
if value2['env_strand'] == '+':
ori_motif=[value2['env_stop'], value2['env_stop']+10, '>', 2, 10, \
color, color, None]
else:
ori_motif=[value2['env_start']-10, value2['env_start'], '<', 2, 10, \
color, color, None]
motifs.append(ori_motif)
#append motif lists to respective annotated gene in dict
gene_dict[key]['motifs'] = motifs
#Set node style
nst_plasmid = NodeStyle()
nst_plasmid['bgcolor'] = 'DarkSeaGreen'
nst_other = NodeStyle()
nst_other = 'AntiqueWhite'
#Now annotate the tree with the motifs
for node in t.traverse():
if node.is_leaf():
if 'plasmid' in node.organism:
node.set_style(nst_plasmid)
else:
node.set_style(nst_other)
seqFace=SeqMotifFace(seq=None, motifs=gene_dict[node.name.split('__')[1]]['motifs'], \
seq_format='blank', gap_format='line')
(t & node.name).add_face(seqFace, 1, 'aligned')
#Create box showing gene percent id
similarity = TextFace(node.pident, fsize=8)
similarity.margin_top = 2
similarity.margin_bottom = 2
similarity.margin_left = 2
similarity.margin_right = 2
#Set box background color based on pident
if node.pident <= 90.0:
similarity.background.color = 'DarkGoldenrod'
elif 90.0 < node.pident <= 95.0:
similarity.background.color = 'ForestGreen'
elif 95.0 <= node.pident:
similarity.background.color = 'YellowGreen'
node.add_face(similarity, column=2, position='aligned')
#Create box showing cluster size
if args.compressed == True:
clust_box = TextFace(node.cluster_size, fsize=8)
clust_box.margin_top = 2
clust_box.margin_bottom = 2
clust_box.margin_left = 2
clust_box.margin_right = 2
node.add_face(clust_box, column=3, position='aligned')
#Return the annotated tree
return t
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
MESSAGE("Setting tree style")
tree_style = TreeStyle()
tree_style.show_leaf_name = False
tree_style.show_branch_length = False
tree_style.draw_guiding_lines = True
tree_style.complete_branch_lines_when_necessary = True
tree_style.legend_position = 1
MESSAGE("Setting legend with condition numbers and colors")
for condi_i in sorted(condi_color_dic.keys()):
tf = TextFace("Condition " + condi_i)
tf.background.color = condi_color_dic[condi_i]
tf.margin_right = 2
tf.margin_top = 1
tf.margin_left = 2
tf.margin_bottom = 1
tf.border.width = 1
tree_style.legend.add_face(tf, column=1)
if add_transition:
MESSAGE("Setting transition style")
tf = TextFace("Transition -> x")
tf.background.color = "white"
tf.margin_right = 2
tf.margin_top = 1
tf.margin_left = 2
tf.margin_bottom = 1
tf.border.color = "red"
tf.border.width = 2
tree_style.legend.add_face(tf, column=1)
