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 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 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 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 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 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 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 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_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 nodeLayoutFunc(node):
taxid = int(node.name)
if taxid in taxidsToKeep:
taxGroupName = ncbiTaxa.get_taxid_translator(
[taxid]
)[taxid] # There has to be an easier way to look up names...
row = None
rangeRows = None
print(len(ranges))
if (len(ranges) == 1):
row = df[(df['ExplanatoryVar'] == var)
& (df['TaxGroup'] == taxid) &
(df['Range'] == ranges[0])]
assert (len(row) == len(ranges))
elif len(ranges) > 1:
row = df[(df['ExplanatoryVar'] == var)
& (df['TaxGroup'] == taxid) &
(df['Range'] == 0)]
assert (len(row) == 1)
rangeRows = df[(df['ExplanatoryVar'] == var)
& (df['TaxGroup'] == taxid) &
(df['Range'].isin(set(ranges)))]
else:
assert (False)
overallPval = float(row['Pvalue'].values[0])
name = TextFace("%s" % taxGroupName,
fsize=baseFontSize * 2.5)
name.tight_text = True
name.margin_left = 20
name.margin_right = 0
name.margin_top = 40
name.margin_bottom = 12
faces.add_face_to_node(name, node, column=0)
#print(rangeRows)
# For each range to be included in this plot, add a bar
for rangeId in ranges:
#print("rangeId = %s" % (rangeId))
rowForThisRange = None
if len(ranges) == 1:
rowForThisRange = row
else:
rowForThisRange = rangeRows[rangeRows['Range'] ==
rangeId]
assert (len(rowForThisRange) == 1)
# Extract p-value and "effect-size" (signed R^2)
effectSize = float(
rowForThisRange['EffectSize'].values[0])
pval = float(rowForThisRange['Pvalue'].values[0])
# Set bar-graph color and significance markers
barColor = ""
significanceMarker = ""
if (pval < significanceLevel):
significanceMarker = " %s" % unichr(0x2731)
if effectSize < 0:
barColor = "#1133ff"
else:
barColor = "#ff3311"
else: # not significant
if effectSize < 0:
barColor = "#b0b0f0"
else:
barColor = "#ccb090"
# Add the minus sign if needed
signChar = ""
if effectSize < 0:
signChar = unichr(
0x2212
) # minus sign (more legible than a hypen...)
v = RectFace(width=abs(effectSize) * barScale,
height=baseFontSize * 3.5,
fgcolor=barColor,
bgcolor=barColor,
label={
"text":
"%s%.2g %s" %
(signChar, abs(effectSize),
significanceMarker),
"fontsize":
baseFontSize * 1.8,
"color":
"black"
})
#v.rotation = -90
v.margin_top = 1
v.margin_left = 30
v.margin_right = 8
v.margin_bottom = 12
faces.add_face_to_node(v, node, column=0)
details = TextFace(
"N=%d" % row['NumSpecies'], fsize=baseFontSize *
1.5) #, fsize=baseFontSize) #, fstyle="italic")
details.background.color = "#dfdfdf"
details.margin_left = 6
details.margin_right = 20
#details.margin_top=5
#details.margin_bottom=0
faces.add_face_to_node(details, node, column=1)
nstyle = NodeStyle()
nstyle["size"] = 0
node.set_style(nstyle)
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)
t = Tree(t[0])
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)
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_heat_tree_V1(taxid2n,
tree_file,
genes,
taxid2st=False,
leaf_label_conversion_dico=False):
'''
Plot heatmap next to a tree. The order of the heatmap **MUST** be the same,
as order of the leafs on the tree. The tree must be in the Newick format. If
*output_file* is specified, then heat-tree will be rendered as a PNG,
otherwise interactive browser will pop-up with your heat-tree.
TODO ajouter en option la possibilite d'ajouter en option la valeur dans la cellule
Parameters
----------
tree_file: str
Path to the tree file in Newick format. The leaf node labels should
be the same as as row names in the heatmap file. E.g. row1, row2.
output_file: str, optional
If specified the heat-tree will be rendered in that file as a PNG image,
otherwise interactive browser will pop-up. **N.B.** program will wait
for you to exit the browser before continuing.
'''
t1 = Tree(tree_file)
tss = TreeStyle()
#t.populate(8)
# Calculate the midpoint node
R = t1.get_midpoint_outgroup()
# and set it as tree outgroup
t1.set_outgroup(R) # To operate with numbers efficiently
import matplotlib.cm as cm
from matplotlib.colors import rgb2hex
import matplotlib as mpl
norm = mpl.colors.Normalize(vmin=0.8, vmax=1) # map2count[map[0]][0]
cmap_blue = cm.Blues
m2 = cm.ScalarMappable(norm=norm, cmap=cmap_blue)
leaf_number = 0
for lf in t1.iter_leaves():
leaf_number += 1
lf.branch_vertical_margin = 0
try:
data = taxid2n[str(lf.name)]
except:
data = [0]
try:
st = taxid2st[lf.name]
except:
st = False
'''
if "taxon2accession_list" not in locals():
from chlamdb.biosqldb import manipulate_biosqldb
server, db = manipulate_biosqldb.load_db("k_cosson_05_16")
sql = 'select taxon_id, accession from bioentry where biodatabase_id=104'
data_tax = server.adaptor.execute_and_fetchall(sql,)
taxon2accession_list = {}
for i in data_tax:
if i[0] not in taxon2accession_list:
taxon2accession_list[i[0]] = [i[1]]
else:
taxon2accession_list[i[0]].append(i[1])
else:
for taxon in taxon2accession_list:
if lf.name in taxon2accession_list[taxon]:
for accession in taxon2accession_list[taxon]:
print lf.name, accession
try:
st = taxid2st[accession]
data = taxid2n[accession]
print 'st ok!!', st
break
except:
continue
'''
if accession2description:
try:
lf.name = accession2description[lf.name]
except:
pass
if st:
lf.name = lf.name + ' (' + st + ')'
else:
pass
for col, value in enumerate(data):
if leaf_number == 1:
n = TextFace('%s' % (genes[col]), fsize=6)
n.vt_align = 2
n.hz_align = 2
n.rotation = 270
n.margin_top = 0
n.margin_right = 0
n.margin_left = 4
n.margin_bottom = 0
n.inner_background.color = "white"
n.opacity = 1.
tss.aligned_header.add_face(n, col)
#lf.add_face(n, col, position="aligned")
if value > 0:
n = TextFace(' ')
n.margin_top = 0
n.margin_right = 0
n.margin_left = 0
n.margin_bottom = 0
n.inner_background.color = rgb2hex(m2.to_rgba(
float(value))) #'#140718' #"#81BEF7"
n.opacity = 1.
lf.add_face(n, col, position="aligned")
else:
n = TextFace(' ')
n.margin_top = 0
n.margin_right = 0
n.margin_left = 0
n.margin_bottom = 0
n.inner_background.color = "white"
n.opacity = 1.
lf.add_face(n, col, position="aligned")
return t1, leaf_number, tss
def plot_heat_tree(tree_file,
biodb="chlamydia_04_16",
exclude_outgroup=False,
bw_scale=True):
from chlamdb.biosqldb import manipulate_biosqldb
import matplotlib.cm as cm
from matplotlib.colors import rgb2hex
import matplotlib as mpl
server, db = manipulate_biosqldb.load_db(biodb)
sql_biodatabase_id = 'select biodatabase_id from biodatabase where name="%s"' % biodb
db_id = server.adaptor.execute_and_fetchall(sql_biodatabase_id, )[0][0]
if type(tree_file) == str:
t1 = Tree(tree_file)
try:
R = t1.get_midpoint_outgroup()
#print 'root', R
# and set it as tree outgroup
t1.set_outgroup(R)
except:
pass
elif isinstance(tree_file, Tree):
t1 = tree_file
else:
IOError('Unkown tree format')
tss = TreeStyle()
tss.draw_guiding_lines = True
tss.guiding_lines_color = "gray"
tss.show_leaf_name = False
#print "tree", t1
sql1 = 'select taxon_id, description from bioentry where biodatabase_id=%s and description not like "%%%%plasmid%%%%"' % db_id
sql2 = 'select t2.taxon_id, t1.GC from genomes_info_%s as t1 inner join bioentry as t2 ' \
' on t1.accession=t2.accession where t2.biodatabase_id=%s and t1.description not like "%%%%plasmid%%%%";' % (biodb, db_id)
sql3 = 'select t2.taxon_id, t1.genome_size from genomes_info_%s as t1 ' \
' inner join bioentry as t2 on t1.accession=t2.accession ' \
' where t2.biodatabase_id=%s and t1.description not like "%%%%plasmid%%%%";' % (biodb, db_id)
sql4 = 'select t2.taxon_id,percent_non_coding from genomes_info_%s as t1 ' \
' inner join bioentry as t2 on t1.accession=t2.accession ' \
' where t2.biodatabase_id=%s and t1.description not like "%%%%plasmid%%%%";' % (biodb, db_id)
sql_checkm_completeness = 'select taxon_id, completeness from custom_tables.checkm_%s;' % biodb
sql_checkm_contamination = 'select taxon_id,contamination from custom_tables.checkm_%s;' % biodb
try:
taxon_id2completeness = manipulate_biosqldb.to_dict(
server.adaptor.execute_and_fetchall(sql_checkm_completeness))
taxon_id2contamination = manipulate_biosqldb.to_dict(
server.adaptor.execute_and_fetchall(sql_checkm_contamination))
except:
taxon_id2completeness = False
#taxon2description = manipulate_biosqldb.to_dict(server.adaptor.execute_and_fetchall(sql1,))
taxon2description = manipulate_biosqldb.taxon_id2genome_description(
server, biodb, filter_names=True)
taxon2gc = manipulate_biosqldb.to_dict(
server.adaptor.execute_and_fetchall(sql2, ))
taxon2genome_size = manipulate_biosqldb.to_dict(
server.adaptor.execute_and_fetchall(sql3, ))
taxon2coding_density = manipulate_biosqldb.to_dict(
server.adaptor.execute_and_fetchall(sql4, ))
my_taxons = [lf.name for lf in t1.iter_leaves()]
# Calculate the midpoint node
if exclude_outgroup:
excluded = str(list(t1.iter_leaves())[0].name)
my_taxons.pop(my_taxons.index(excluded))
genome_sizes = [float(taxon2genome_size[i]) for i in my_taxons]
gc_list = [float(taxon2gc[i]) for i in my_taxons]
fraction_list = [float(taxon2coding_density[i]) for i in my_taxons]
value = 1
max_genome_size = max(genome_sizes) #3424182#
max_gc = max(gc_list) #48.23
cmap = cm.YlGnBu #YlOrRd#OrRd
norm = mpl.colors.Normalize(vmin=min(genome_sizes) - 100000,
vmax=max(genome_sizes))
m1 = cm.ScalarMappable(norm=norm, cmap=cmap)
norm = mpl.colors.Normalize(vmin=min(gc_list), vmax=max(gc_list))
m2 = cm.ScalarMappable(norm=norm, cmap=cmap)
norm = mpl.colors.Normalize(vmin=min(fraction_list),
vmax=max(fraction_list))
m3 = cm.ScalarMappable(norm=norm, cmap=cmap)
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:
n = TextFace('Size (Mbp)')
n.rotation = -25
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, 3, position="aligned")
tss.aligned_header.add_face(n, 3)
n = TextFace('GC (%)')
n.rotation = -25
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, 5, position="aligned")
tss.aligned_header.add_face(n, 5)
n = TextFace('')
#lf.add_face(n, 2, position="aligned")
tss.aligned_header.add_face(n, 2)
#lf.add_face(n, 4, position="aligned")
tss.aligned_header.add_face(n, 4)
n = TextFace('Non coding (%)')
n.margin_top = 1
n.margin_right = 1
n.margin_left = 20
n.margin_bottom = 1
n.inner_background.color = "white"
n.opacity = 1.
n.rotation = -25
#lf.add_face(n, 7, position="aligned")
tss.aligned_header.add_face(n, 7)
n = TextFace('')
#lf.add_face(n, 6, position="aligned")
tss.aligned_header.add_face(n, 6)
if taxon_id2completeness:
n = TextFace('Completeness (%)')
n.margin_top = 1
n.margin_right = 1
n.margin_left = 20
n.margin_bottom = 1
n.inner_background.color = "white"
n.opacity = 1.
n.rotation = -25
#lf.add_face(n, 7, position="aligned")
tss.aligned_header.add_face(n, 9)
n = TextFace('')
#lf.add_face(n, 6, position="aligned")
tss.aligned_header.add_face(n, 8)
n = TextFace('Contamination (%)')
n.margin_top = 1
n.margin_right = 1
n.margin_left = 20
n.margin_bottom = 1
n.inner_background.color = "white"
n.opacity = 1.
n.rotation = -25
#lf.add_face(n, 7, position="aligned")
tss.aligned_header.add_face(n, 11)
n = TextFace('')
#lf.add_face(n, 6, position="aligned")
tss.aligned_header.add_face(n, 10)
value += 1
#print '------ %s' % lf.name
if exclude_outgroup and i == 0:
lf.name = taxon2description[lf.name]
#print '#######################'
continue
n = TextFace(
' %s ' %
str(round(taxon2genome_size[lf.name] / float(1000000), 2)))
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")
#if max_genome_size > 3424182:
# max_genome_size = 3424182
fraction_biggest = (float(taxon2genome_size[lf.name]) /
max_genome_size) * 100
fraction_rest = 100 - fraction_biggest
if taxon2description[lf.name] == 'Rhabdochlamydia helveticae T3358':
col = '#fc8d59'
else:
if not bw_scale:
col = rgb2hex(m1.to_rgba(float(
taxon2genome_size[lf.name]))) # 'grey'
else:
col = '#fc8d59'
b = StackedBarFace([fraction_biggest, fraction_rest],
width=100,
height=9,
colors=[col, 'white'])
b.rotation = 0
b.inner_border.color = "black"
b.inner_border.width = 0
b.margin_right = 15
b.margin_left = 0
lf.add_face(b, 3, position="aligned")
fraction_biggest = (float(taxon2gc[lf.name]) / max_gc) * 100
fraction_rest = 100 - fraction_biggest
if taxon2description[lf.name] == 'Rhabdochlamydia helveticae T3358':
col = '#91bfdb'
else:
if not bw_scale:
col = rgb2hex(m2.to_rgba(float(taxon2gc[lf.name])))
else:
col = '#91bfdb'
b = StackedBarFace([fraction_biggest, fraction_rest],
width=100,
height=9,
colors=[col, 'white'])
b.rotation = 0
b.inner_border.color = "black"
b.inner_border.width = 0
b.margin_left = 0
b.margin_right = 15
lf.add_face(b, 5, position="aligned")
n = TextFace(' %s ' % str(round(float(taxon2gc[lf.name]), 2)))
n.margin_top = 1
n.margin_right = 0
n.margin_left = 0
n.margin_bottom = 1
n.fsize = 7
n.inner_background.color = "white"
n.opacity = 1.
lf.add_face(n, 4, position="aligned")
if taxon2description[lf.name] == 'Rhabdochlamydia helveticae T3358':
col = '#99d594'
else:
if not bw_scale:
col = rgb2hex(m3.to_rgba(float(taxon2coding_density[lf.name])))
else:
col = '#99d594'
n = TextFace(' %s ' % str(float(taxon2coding_density[lf.name])))
n.margin_top = 1
n.margin_right = 0
n.margin_left = 0
n.margin_right = 0
n.margin_bottom = 1
n.fsize = 7
n.inner_background.color = "white"
n.opacity = 1.
lf.add_face(n, 6, position="aligned")
fraction = (float(taxon2coding_density[lf.name]) /
max(taxon2coding_density.values())) * 100
fraction_rest = ((max(taxon2coding_density.values()) -
taxon2coding_density[lf.name]) /
float(max(taxon2coding_density.values()))) * 100
#print 'fraction, rest', fraction, fraction_rest
b = StackedBarFace(
[fraction, fraction_rest],
width=100,
height=9,
colors=[col, 'white'
]) # 1-round(float(taxon2coding_density[lf.name]), 2)
b.rotation = 0
b.margin_right = 1
b.inner_border.color = "black"
b.inner_border.width = 0
b.margin_left = 5
lf.add_face(b, 7, position="aligned")
if taxon_id2completeness:
n = TextFace(' %s ' % str(float(taxon_id2completeness[lf.name])))
n.margin_top = 1
n.margin_right = 0
n.margin_left = 0
n.margin_right = 0
n.margin_bottom = 1
n.fsize = 7
n.inner_background.color = "white"
n.opacity = 1.
lf.add_face(n, 8, position="aligned")
fraction = float(taxon_id2completeness[lf.name])
fraction_rest = 100 - fraction
#print 'fraction, rest', fraction, fraction_rest
b = StackedBarFace(
[fraction, fraction_rest],
width=100,
height=9,
colors=["#d7191c", 'white'
]) # 1-round(float(taxon2coding_density[lf.name]), 2)
b.rotation = 0
b.margin_right = 1
b.inner_border.color = "black"
b.inner_border.width = 0
b.margin_left = 5
lf.add_face(b, 9, position="aligned")
n = TextFace(' %s ' % str(float(taxon_id2contamination[lf.name])))
n.margin_top = 1
n.margin_right = 0
n.margin_left = 0
n.margin_right = 0
n.margin_bottom = 1
n.fsize = 7
n.inner_background.color = "white"
n.opacity = 1.
lf.add_face(n, 10, position="aligned")
fraction = float(taxon_id2contamination[lf.name])
fraction_rest = 100 - fraction
#print 'fraction, rest', fraction, fraction_rest
b = StackedBarFace(
[fraction, fraction_rest],
width=100,
height=9,
colors=["black", 'white'
]) # 1-round(float(taxon2coding_density[lf.name]), 2)
b.rotation = 0
b.margin_right = 1
b.inner_border.color = "black"
b.inner_border.width = 0
b.margin_left = 5
lf.add_face(b, 11, position="aligned")
#lf.name = taxon2description[lf.name]
n = TextFace(taxon2description[lf.name],
fgcolor="black",
fsize=9,
fstyle='italic')
n.margin_right = 30
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_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
nstyle["size"] = 1
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
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_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 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_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 plot_tree_text_metadata(tree_file, header2taxon2text, ordered_header_list,
biodb):
from chlamdb.biosqldb import manipulate_biosqldb
server, db = manipulate_biosqldb.load_db(biodb)
t1 = Tree(tree_file)
taxon2description = manipulate_biosqldb.taxon_id2genome_description(
server, biodb, filter_names=True)
# Calculate the midpoint node
R = t1.get_midpoint_outgroup()
# and set it as tree outgroup
t1.set_outgroup(R)
tss = TreeStyle()
tss.draw_guiding_lines = True
tss.guiding_lines_color = "gray"
tss.show_leaf_name = False
for i, leaf in enumerate(t1.iter_leaves()):
# first leaf, add headers
if i == 0:
for column, header in enumerate(ordered_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, column)
for column, header in enumerate(ordered_header_list):
text = header2taxon2text[header][int(leaf.name)]
n = TextFace('%s' % text)
n.margin_top = 1
n.margin_right = 1
n.margin_left = 5
n.margin_bottom = 1
n.inner_background.color = "white"
n.opacity = 1.
#n.rotation = 270
leaf.add_face(n, column + 1, position="aligned")
# rename leaf (taxon_id => description)
n = TextFace(taxon2description[leaf.name],
fgcolor="black",
fsize=12,
fstyle='italic')
leaf.add_face(n, 0)
for n in t1.traverse():
# rename leaf
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 custom_layout(node):
ncbi=connect_ncbitaxa()
if node.is_leaf():
total_name = (node.name)
if not total_name or total_name == "":
sys.stderr.write("Name of node is null or empty when creating custom layout.\n")
return
#seq_name = (total_name.split('.', 1)[-1])
seq_name = (total_name.split('|')[1])
other_info = (total_name.split('|')[2])
aligned_name_face = TextFace(seq_name, fgcolor='brown', fsize=11)
aligned_name_face.margin_top = 0
aligned_name_face.margin_bottom = 0
aligned_name_face.margin_left = 5
add_face_to_node(aligned_name_face, node, column=2, position='aligned')
#node.name=(node.name.split('|')[0])
node_name = node.name.split('|')[0]
if not node_name or node_name.strip() == "":
sys.stderr.write("Node name is null or empty when creating custom layout.\n")
return
name2taxid=ncbi.get_name_translator([node_name])
taxid=name2taxid[node_name]
lin = ncbi.get_lineage(int(taxid[0]))
if int('7742') in lin:
N = TextFace('vertebrata', fsize=11, fgcolor="red")
N.margin_left = 5
N.background.color = "Linen"
add_face_to_node(N, node, column=3, position = 'aligned')
if int('6040') in lin:
N = TextFace('porifera', fsize=11, fgcolor="green")
N.margin_left = 5
N.background.color = "Linen"
add_face_to_node(N, node, column=3, position = 'aligned')
if int('6073') in lin:
N = TextFace('cnidario', fsize=11, fgcolor="orange")
N.margin_left = 5
N.background.color = "Linen"
add_face_to_node(N, node, column=3, position = 'aligned')
if int('33317') in lin:
N = TextFace('protostomia', fsize=11, fgcolor="blue")
N.margin_left = 5
N.background.color = "Linen"
add_face_to_node(N, node, column=3, position = 'aligned')
if int('10197') in lin:
N = TextFace('Ctenophora', fsize=11, fgcolor="indigo")
N.margin_left = 5
N.background.color = "Linen"
add_face_to_node(N, node, column=3, position = 'aligned')
if int('10226') in lin:
N = TextFace('Ctenophora', fsize=11, fgcolor="sienna")
N.margin_left = 5
N.background.color = "Linen"
add_face_to_node(N, node, column=3, position = 'aligned')
if int('6157') in lin:
N = TextFace('Platyhelminthes', fsize=11, fgcolor="olive")
N.margin_left = 5
N.background.color = "Linen"
add_face_to_node(N, node, column=3, position = 'aligned')
if int('7735') in lin:
N = TextFace('Cephalochordata', fsize=11, fgcolor="skyblue")
N.margin_left = 5
N.background.color = "Linen"
add_face_to_node(N, node, column=3, position = 'aligned')
tax, seqs_info = other_info.split('.', 1)
try:
tax = int(tax)
except:
tax = tax
if tax in lin:
aligned_name_face = TextFace(seqs_info, fgcolor='grey', fsize=11)
aligned_name_face.margin_top = 0
aligned_name_face.margin_bottom = 0
aligned_name_face.margin_left = 5
add_face_to_node(aligned_name_face, node, column=4, position='aligned')
else:
aligned_name_face = TextFace(other_info, fgcolor='red', fsize=11)
aligned_name_face.margin_top = 0
aligned_name_face.margin_bottom = 0
aligned_name_face.margin_left = 5
add_face_to_node(aligned_name_face, node, column=4, position='aligned')
seqFace = SeqMotifFace(node.sequence, gap_format="blank")
add_face_to_node(seqFace, node, column=5, position="aligned")
node.img_style['size'] = 0
#try:
# g_sym=gene_sym[sci_name]
# predNameFace = faces.TextFace(g_sym,fgcolor = "navy" , fsize=28)
# add_face_to_node(predNameFace, node, column=3, position="branch-right" )
#except:
# predNameFace = faces.TextFace(' ',fgcolor="navy", fsize=28)
# add_face_to_node(predNameFace, node, column=3, position="branch-right")
if node_name.startswith("H**o"):
# Add an static face that handles the node name
N = TextFace(node_name, fsize=11, fgcolor="red")
add_face_to_node(N, node, column=0, position = 'branch-right')
elif node_name.startswith("Spongilla"):
N = TextFace(node_name, fsize=11, fgcolor="green")
add_face_to_node(N, node, column=0, position = 'branch-right')
elif node_name.startswith("Sycon"):
N = TextFace(node_name, fsize=11, fgcolor="green")
add_face_to_node(N, node, column=0, position = 'branch-right')
elif node_name.startswith("Amphimedon"):
N = TextFace(node_name, fsize=11, fgcolor="green")
add_face_to_node(N, node, column=0, position = 'branch-right')
elif node_name.startswith("Oscarella"):
N = TextFace(node_name, fsize=11, fgcolor="green")
add_face_to_node(N, node, column=0, position = 'branch-right')
elif node_name.startswith("Gallus"):
N = TextFace(node_name, fsize=11, fgcolor="red")
add_face_to_node(N, node, column=0, position = 'branch-right')
elif node_name.startswith("Branchiostoma"):
N = TextFace(node_name, fsize=11, fgcolor="red")
add_face_to_node(N, node, column=0, position = 'branch-right')
elif node_name.startswith("Trichoplax"):
N = TextFace(node_name, fsize=11, fgcolor="orange")
add_face_to_node(N, node, column=0, position = 'branch-right')
elif node_name.startswith("Nematostella"):
N = TextFace(node_name, fsize=11, fgcolor="orange")
add_face_to_node(N, node, column=0, position = 'branch-right')
elif node_name.startswith("Hydra"):
N = TextFace(node_name, fsize=11, fgcolor="orange")
add_face_to_node(N, node, column=0, position = 'branch-right')
elif node_name.startswith("Drosophila"):
N = TextFace(node_name, fsize=11, fgcolor="blue")
add_face_to_node(N, node, column=0, position = 'branch-right')
elif node_name.startswith("Crassostrea"):
N = TextFace(node_name, fsize=11, fgcolor="blue")
add_face_to_node(N, node, column=0, position = 'branch-right')
else:
name_face = TextFace(node_name, fgcolor='#333333', fsize=11)
name_face.margin_top = 0
name_face.margin_bottom = 0
add_face_to_node(name_face, node, column=0, position='branch-right')
else:
node.img_style['size'] = 3
node.img_style['shape'] = 'square'
if node.name:
name_face = TextFace(node.name, fgcolor='grey', fsize=10)
name_face.margin_bottom = 1
add_face_to_node(name_face, node, column=0, position='branch-top')
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')
return
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 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
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_heatmap_tree_locus(biodb,
tree_file,
taxid2count,
taxid2identity=False,
taxid2locus=False,
reference_taxon=False,
n_paralogs_barplot=False):
'''
plot tree and associated heatmap with count of homolgs
optional:
- add identity of closest homolog
- add locus tag of closest homolog
'''
from chlamdb.biosqldb import manipulate_biosqldb
server, db = manipulate_biosqldb.load_db(biodb)
taxid2organism = manipulate_biosqldb.taxon_id2genome_description(
server, biodb, True)
t1 = Tree(tree_file)
ts = TreeStyle()
ts.draw_guiding_lines = True
ts.guiding_lines_color = "gray"
# Calculate the midpoint node
R = t1.get_midpoint_outgroup()
# and set it as tree outgroup
t1.set_outgroup(R)
leaf_number = 0
for lf in t1.iter_leaves():
if str(lf.name) not in taxid2count:
taxid2count[str(lf.name)] = 0
max_count = max([taxid2count[str(lf.name)] for lf in t1.iter_leaves()])
for i, lf in enumerate(t1.iter_leaves()):
# top leaf, add header
if i == 0:
n = TextFace('Number of homologs')
n.margin_top = 1
n.margin_right = 1
n.margin_left = 20
n.margin_bottom = 1
n.inner_background.color = "white"
n.opacity = 1.
n.rotation = -25
#lf.add_face(n, 7, position="aligned")
ts.aligned_header.add_face(n, 1)
if taxid2identity:
n = TextFace('Protein identity')
n.margin_top = 1
n.margin_right = 1
n.margin_left = 20
n.margin_bottom = 1
n.inner_background.color = "white"
n.opacity = 1.
n.rotation = -25
#lf.add_face(n, 7, position="aligned")
ts.aligned_header.add_face(n, 2)
if taxid2locus:
n = TextFace('Locus tag')
n.margin_top = 1
n.margin_right = 1
n.margin_left = 20
n.margin_bottom = 1
n.inner_background.color = "white"
n.opacity = 1.
n.rotation = -25
#lf.add_face(n, 7, position="aligned")
ts.aligned_header.add_face(n, 3)
leaf_number += 1
lf.branch_vertical_margin = 0
data = [taxid2count[str(lf.name)]]
# possibility to add one or more columns
for col, value in enumerate(data):
col_index = col
if value > 0:
n = TextFace(' %s ' % str(value))
n.margin_top = 2
n.margin_right = 2
if col == 0:
n.margin_left = 20
else:
n.margin_left = 2
n.margin_bottom = 2
n.inner_background.color = "white" # #81BEF7
n.opacity = 1.
lf.add_face(n, col, position="aligned")
else:
n = TextFace(' %s ' % str(value))
n.margin_top = 2
n.margin_right = 2
if col == 0:
n.margin_left = 20
else:
n.margin_left = 2
n.margin_bottom = 2
n.inner_background.color = "white"
n.opacity = 1.
lf.add_face(n, col, position="aligned")
# optionally indicate number of paralogs as a barplot
if n_paralogs_barplot:
col_index += 1
percent = (float(value) / max_count) * 100
n = StackedBarFace([percent, 100 - percent],
width=150,
height=18,
colors=['#6699ff', 'white'],
line_color='white')
n.rotation = 0
n.inner_border.color = "white"
n.inner_border.width = 0
n.margin_right = 15
n.margin_left = 0
lf.add_face(n, col + 1, position="aligned")
# optionally add additionnal column with identity
if taxid2identity:
import matplotlib.cm as cm
from matplotlib.colors import rgb2hex
import matplotlib as mpl
norm = mpl.colors.Normalize(vmin=0, vmax=100)
cmap = cm.OrRd
m = cm.ScalarMappable(norm=norm, cmap=cmap)
try:
if round(taxid2identity[str(lf.name)], 2) != 100:
value = "%.2f" % round(taxid2identity[str(lf.name)], 2)
else:
value = "%.1f" % round(taxid2identity[str(lf.name)], 2)
except:
value = '-'
if str(lf.name) == str(reference_taxon):
value = ' '
n = TextFace(' %s ' % value)
n.margin_top = 2
n.margin_right = 2
n.margin_left = 20
n.margin_bottom = 2
if not value.isspace() and value is not '-':
n.inner_background.color = rgb2hex(m.to_rgba(float(value)))
if float(value) > 82:
n.fgcolor = 'white'
n.opacity = 1.
if str(lf.name) == str(reference_taxon):
n.inner_background.color = '#800000'
lf.add_face(n, col_index + 1, position="aligned")
# optionaly add column with locus name
if taxid2locus:
try:
value = str(taxid2locus[str(lf.name)])
except:
value = '-'
n = TextFace(' %s ' % value)
n.margin_top = 2
n.margin_right = 2
n.margin_left = 2
n.margin_bottom = 2
if str(lf.name) != str(reference_taxon):
n.inner_background.color = "white"
else:
n.fgcolor = '#ff0000'
n.inner_background.color = "white"
n.opacity = 1.
lf.add_face(n, col_index + 2, position="aligned")
lf.name = taxid2organism[str(lf.name)]
return t1, leaf_number, ts