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 _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 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(biodb, taxid2n, tree_file):
'''
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.
Parameters
----------
heatmap_file: str
Path to the heatmap file. The first row must have '#Names' as first
element of the header.
e.g. #Names, A, B, C, D
row1, 2, 4, 0, 4
row2, 4, 6, 2, -1
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.
'''
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)
# 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():
leaf_number += 1
lf.branch_vertical_margin = 0
try:
data = [taxid2n[str(lf.name)]]
except:
data = [0]
#print 'taxon', int(lf.name)
lf.name = taxid2organism[int(lf.name)]
for col, value in enumerate(data):
if value > 0:
n = TextFace(' %s ' % str(value))
n.margin_top = 2
n.margin_right = 2
n.margin_left = 2
n.margin_bottom = 2
n.inner_background.color = "#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
n.margin_left = 2
n.margin_bottom = 2
n.inner_background.color = "white"
n.opacity = 1.
lf.add_face(n, col, position="aligned")
return t1, leaf_number
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
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
mark3.margin_right = 20
mark3.margin_left = 5
mark3.margin_bottom = 2
ts.legend.add_face(mark3, column=1)
ts.margin_left = 20
ts.margin_right = 20
ts.margin_top = 10
ts.margin_bottom = 10
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 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
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 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_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
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 run(args):
if args.text_mode:
from ete3 import Tree
for tindex, tfile in enumerate(args.src_tree_iterator):
#print tfile
if args.raxml:
nw = re.sub(":(\d+\.\d+)\[(\d+)\]", ":\\1[&&NHX:support=\\2]",
open(tfile).read())
t = Tree(nw)
else:
t = Tree(tfile)
print(
t.get_ascii(show_internal=args.show_internal_names,
attributes=args.show_attributes))
return
import random
import re
import colorsys
from collections import defaultdict
from ete3 import (Tree, PhyloTree, TextFace, RectFace, faces, TreeStyle,
add_face_to_node, random_color)
global FACES
if args.face:
FACES = parse_faces(args.face)
else:
FACES = []
# VISUALIZATION
ts = TreeStyle()
ts.mode = args.mode
ts.show_leaf_name = True
ts.tree_width = args.tree_width
for f in FACES:
if f["value"] == "@name":
ts.show_leaf_name = False
break
if args.as_ncbi:
ts.show_leaf_name = False
FACES.extend(
parse_faces([
'value:@sci_name, size:10, fstyle:italic',
'value:@taxid, color:grey, size:6, format:" - %s"',
'value:@sci_name, color:steelblue, size:7, pos:b-top, nodetype:internal',
'value:@rank, color:indianred, size:6, pos:b-bottom, nodetype:internal',
]))
if args.alg:
FACES.extend(
parse_faces([
'value:@sequence, size:10, pos:aligned, ftype:%s' %
args.alg_type
]))
if args.heatmap:
FACES.extend(
parse_faces(['value:@name, size:10, pos:aligned, ftype:heatmap']))
if args.bubbles:
for bubble in args.bubbles:
FACES.extend(
parse_faces([
'value:@%s, pos:float, ftype:bubble, opacity:0.4' % bubble,
]))
ts.branch_vertical_margin = args.branch_separation
if args.show_support:
ts.show_branch_support = True
if args.show_branch_length:
ts.show_branch_length = True
if args.force_topology:
ts.force_topology = True
ts.layout_fn = lambda x: None
for tindex, tfile in enumerate(args.src_tree_iterator):
#print tfile
if args.raxml:
nw = re.sub(":(\d+\.\d+)\[(\d+)\]", ":\\1[&&NHX:support=\\2]",
open(tfile).read())
t = PhyloTree(nw)
else:
t = PhyloTree(tfile)
if args.alg:
t.link_to_alignment(args.alg, alg_format=args.alg_format)
if args.heatmap:
DEFAULT_COLOR_SATURATION = 0.3
BASE_LIGHTNESS = 0.7
def gradient_color(value, max_value, saturation=0.5, hue=0.1):
def rgb2hex(rgb):
return '#%02x%02x%02x' % rgb
def hls2hex(h, l, s):
return rgb2hex(
tuple([
int(x * 255) for x in colorsys.hls_to_rgb(h, l, s)
]))
lightness = 1 - (value * BASE_LIGHTNESS) / max_value
return hls2hex(hue, lightness, DEFAULT_COLOR_SATURATION)
heatmap_data = {}
max_value, min_value = None, None
for line in open(args.heatmap):
if line.startswith('#COLNAMES'):
pass
elif line.startswith('#') or not line.strip():
pass
else:
fields = line.split('\t')
name = fields[0].strip()
values = [float(x) if x else None for x in fields[1:]]
maxv = max(values)
minv = min(values)
if max_value is None or maxv > max_value:
max_value = maxv
if min_value is None or minv < min_value:
min_value = minv
heatmap_data[name] = values
heatmap_center_value = 0
heatmap_color_center = "white"
heatmap_color_up = 0.3
heatmap_color_down = 0.7
heatmap_color_missing = "black"
heatmap_max_value = abs(heatmap_center_value - max_value)
heatmap_min_value = abs(heatmap_center_value - min_value)
if heatmap_center_value <= min_value:
heatmap_max_value = heatmap_min_value + heatmap_max_value
else:
heatmap_max_value = max(heatmap_min_value, heatmap_max_value)
# scale the tree
if not args.height:
args.height = None
if not args.width:
args.width = None
f2color = {}
f2last_seed = {}
for node in t.traverse():
node.img_style['size'] = 0
if len(node.children) == 1:
node.img_style['size'] = 2
node.img_style['shape'] = "square"
node.img_style['fgcolor'] = "steelblue"
ftype_pos = defaultdict(int)
for findex, f in enumerate(FACES):
if (f['nodetype'] == 'any'
or (f['nodetype'] == 'leaf' and node.is_leaf()) or
(f['nodetype'] == 'internal' and not node.is_leaf())):
# if node passes face filters
if node_matcher(node, f["filters"]):
if f["value"].startswith("@"):
fvalue = getattr(node, f["value"][1:], None)
else:
fvalue = f["value"]
# if node's attribute has content, generate face
if fvalue is not None:
fsize = f["size"]
fbgcolor = f["bgcolor"]
fcolor = f['color']
if fcolor:
# Parse color options
auto_m = re.search("auto\(([^)]*)\)", fcolor)
if auto_m:
target_attr = auto_m.groups()[0].strip()
if not target_attr:
color_keyattr = f["value"]
else:
color_keyattr = target_attr
color_keyattr = color_keyattr.lstrip('@')
color_bin = getattr(
node, color_keyattr, None)
last_seed = f2last_seed.setdefault(
color_keyattr, random.random())
seed = last_seed + 0.10 + random.uniform(
0.1, 0.2)
f2last_seed[color_keyattr] = seed
fcolor = f2color.setdefault(
color_bin, random_color(h=seed))
if fbgcolor:
# Parse color options
auto_m = re.search("auto\(([^)]*)\)", fbgcolor)
if auto_m:
target_attr = auto_m.groups()[0].strip()
if not target_attr:
color_keyattr = f["value"]
else:
color_keyattr = target_attr
color_keyattr = color_keyattr.lstrip('@')
color_bin = getattr(
node, color_keyattr, None)
last_seed = f2last_seed.setdefault(
color_keyattr, random.random())
seed = last_seed + 0.10 + random.uniform(
0.1, 0.2)
f2last_seed[color_keyattr] = seed
fbgcolor = f2color.setdefault(
color_bin, random_color(h=seed))
if f["ftype"] == "text":
if f.get("format", None):
fvalue = f["format"] % fvalue
F = TextFace(fvalue,
fsize=fsize,
fgcolor=fcolor or "black",
fstyle=f.get('fstyle', None))
elif f["ftype"] == "fullseq":
F = faces.SeqMotifFace(seq=fvalue,
seq_format="seq",
seqtail_format="seq",
height=fsize)
elif f["ftype"] == "compactseq":
F = faces.SeqMotifFace(
seq=fvalue,
seq_format="compactseq",
seqtail_format="compactseq",
height=fsize)
elif f["ftype"] == "blockseq":
F = faces.SeqMotifFace(
seq=fvalue,
seq_format="blockseq",
seqtail_format="blockseq",
height=fsize,
fgcolor=fcolor or "slategrey",
bgcolor=fbgcolor or "slategrey",
scale_factor=1.0)
fbgcolor = None
elif f["ftype"] == "bubble":
try:
v = float(fvalue)
except ValueError:
rad = fsize
else:
rad = fsize * v
F = faces.CircleFace(radius=rad,
style="sphere",
color=fcolor
or "steelblue")
elif f["ftype"] == "heatmap":
if not f['column']:
col = ftype_pos[f["pos"]]
else:
col = f["column"]
for i, value in enumerate(
heatmap_data.get(node.name, [])):
ftype_pos[f["pos"]] += 1
if value is None:
color = heatmap_color_missing
elif value > heatmap_center_value:
color = gradient_color(
abs(heatmap_center_value - value),
heatmap_max_value,
hue=heatmap_color_up)
elif value < heatmap_center_value:
color = gradient_color(
abs(heatmap_center_value - value),
heatmap_max_value,
hue=heatmap_color_down)
else:
color = heatmap_color_center
node.add_face(RectFace(
20, 20, color, color),
position="aligned",
column=col + i)
# Add header
# for i, name in enumerate(header):
# nameF = TextFace(name, fsize=7)
# nameF.rotation = -90
# tree_style.aligned_header.add_face(nameF, column=i)
F = None
elif f["ftype"] == "profile":
# internal profiles?
F = None
elif f["ftype"] == "barchart":
F = None
elif f["ftype"] == "piechart":
F = None
# Add the Face
if F:
F.opacity = f['opacity'] or 1.0
# Set face general attributes
if fbgcolor:
F.background.color = fbgcolor
if not f['column']:
col = ftype_pos[f["pos"]]
ftype_pos[f["pos"]] += 1
else:
col = f["column"]
node.add_face(F, column=col, position=f["pos"])
if args.image:
t.render("t%d.%s" % (tindex, args.image),
tree_style=ts,
w=args.width,
h=args.height,
units=args.size_units)
else:
t.show(None, tree_style=ts)
def ete_layout(node):
"""
Formatting of tree nodes while tree is rendered
:param node: ete node
"""
nstyle = NodeStyle()
nstyle["shape"] = "sphere"
nstyle["size"] = 0
nstyle["vt_line_width"] = 1 # line width of vertical branches
nstyle["hz_line_width"] = 1 # line width of horizontal branches
node.set_style(nstyle)
if 'median_vaf' in node.features and not np.isnan(
node.median_vaf) and logger.isEnabledFor(logging.DEBUG):
# # Creates a sphere face whose size is proportional to the given VAF or CCF of the acquired muts
# cf = CircleFace(radius=node.median_vaf*50, color="RoyalBlue", style="sphere")
# # Let's make the sphere transparent
# cf.opacity = 0.6
# # And place as a float face over the tree
# faces.add_face_to_node(cf, node, column=0, position="float-behind")
vaf_face = TextFace('({:.0%}) '.format(node.median_vaf),
fsize=9,
fgcolor='SlateGray')
vaf_face.opacity = 0.8 # from 0 to 1
faces.add_face_to_node(vaf_face, node, column=2, position="branch-top")
if node.is_root():
# tf_root = TextFace(node.name, fsize=14, fgcolor="Black")
# tf_root.margin_bottom = 25
# faces.add_face_to_node(tf_root, node, column=0, position='branch-right')
return
bt_face_muts = AttrFace("dist",
fsize=11,
fgcolor="Black",
text_prefix=' ',
text_suffix=' ',
formatter='%d')
bt_face_muts.margin_bottom = 2
bt_face_muts.margin_top = 2
faces.add_face_to_node(bt_face_muts, node, column=1, position="branch-top")
if 'drivers' in node.features and len(node.drivers) > 0:
drivers_face = TextFace(node.drivers,
fsize=11,
fstyle='italic',
fgcolor='OrangeRed')
drivers_face.opacity = 0.8 # from 0 to 1
drivers_face.margin_left = 1
drivers_face.margin_right = 1
faces.add_face_to_node(drivers_face,
node,
column=0,
position="branch-top")
# If node is a leaf, add the nodes name and a its scientific name
if node.is_leaf():
# for colors see: http://etetoolkit.org/docs/latest/reference/reference_treeview.html#color-names
leaf_face = AttrFace("name",
fsize=14,
fgcolor="Black",
text_prefix=' ',
text_suffix=' ')
if node.name.startswith('PT') or node.name.startswith(
'Primary'): # primary tumor sample
leaf_face.fgcolor = 'SteelBlue'
elif node.name.startswith('LiM'): # liver met
leaf_face.fgcolor = 'DarkOliveGreen'
elif node.name.startswith('LuM'): # lung met
leaf_face.fgcolor = 'SaddleBrown'
elif node.name.startswith('NoM'): # lymph node met
leaf_face.fgcolor = 'DarkViolet'
elif node.name.startswith('PeM'): # peritoneal met
leaf_face.fgcolor = 'Purple'
elif node.name.startswith('Met') or node.name.startswith(
'M'): # some metastasis
leaf_face.fgcolor = 'Magenta'
elif node.name.startswith('BrM'): # brain met
leaf_face.fgcolor = 'Crimson'
leaf_face.border.type = 0
leaf_face.border.width = 1
leaf_face.margin_bottom = 1
leaf_face.margin_top = 1
faces.add_face_to_node(leaf_face, node,
column=1) # , position="aligned"
else: # inner node
# if not np.isnan(node.support):
# bt_face_conf = AttrFace("support", fsize=12, fgcolor="DimGrey", text_prefix=' ',
# text_suffix=' ', formatter='%0.2f')
if node.confidence is not None and node.confidence != '':
tf_conf = TextFace(node.confidence, fsize=12, fgcolor='DimGrey')
tf_conf.margin_left = 3
tf_conf.margin_right = 3
faces.add_face_to_node(tf_conf,
node,
column=1,
position="branch-bottom")
def run(args):
if args.text_mode:
from ete3 import Tree
for tindex, tfile in enumerate(args.src_tree_iterator):
#print tfile
if args.raxml:
nw = re.sub(":(\d+\.\d+)\[(\d+)\]", ":\\1[&&NHX:support=\\2]", open(tfile).read())
t = Tree(nw)
else:
t = Tree(tfile)
print(t.get_ascii(show_internal=args.show_internal_names,
attributes=args.show_attributes))
return
import random
import re
import colorsys
from collections import defaultdict
from ete3 import (Tree, PhyloTree, TextFace, RectFace, faces, TreeStyle,
add_face_to_node, random_color)
global FACES
if args.face:
FACES = parse_faces(args.face)
else:
FACES = []
# VISUALIZATION
ts = TreeStyle()
ts.mode = args.mode
ts.show_leaf_name = True
ts.tree_width = args.tree_width
for f in FACES:
if f["value"] == "@name":
ts.show_leaf_name = False
break
if args.as_ncbi:
ts.show_leaf_name = False
FACES.extend(parse_faces(
['value:@sci_name, size:10, fstyle:italic',
'value:@taxid, color:grey, size:6, format:" - %s"',
'value:@sci_name, color:steelblue, size:7, pos:b-top, nodetype:internal',
'value:@rank, color:indianred, size:6, pos:b-bottom, nodetype:internal',
]))
if args.alg:
FACES.extend(parse_faces(
['value:@sequence, size:10, pos:aligned, ftype:%s' %args.alg_type]
))
if args.heatmap:
FACES.extend(parse_faces(
['value:@name, size:10, pos:aligned, ftype:heatmap']
))
if args.bubbles:
for bubble in args.bubbles:
FACES.extend(parse_faces(
['value:@%s, pos:float, ftype:bubble, opacity:0.4' %bubble,
]))
ts.branch_vertical_margin = args.branch_separation
if args.show_support:
ts.show_branch_support = True
if args.show_branch_length:
ts.show_branch_length = True
if args.force_topology:
ts.force_topology = True
ts.layout_fn = lambda x: None
for tindex, tfile in enumerate(args.src_tree_iterator):
#print tfile
if args.raxml:
nw = re.sub(":(\d+\.\d+)\[(\d+)\]", ":\\1[&&NHX:support=\\2]", open(tfile).read())
t = PhyloTree(nw)
else:
t = PhyloTree(tfile)
if args.alg:
t.link_to_alignment(args.alg, alg_format=args.alg_format)
if args.heatmap:
DEFAULT_COLOR_SATURATION = 0.3
BASE_LIGHTNESS = 0.7
def gradient_color(value, max_value, saturation=0.5, hue=0.1):
def rgb2hex(rgb):
return '#%02x%02x%02x' % rgb
def hls2hex(h, l, s):
return rgb2hex( tuple([int(x*255) for x in colorsys.hls_to_rgb(h, l, s)]))
lightness = 1 - (value * BASE_LIGHTNESS) / max_value
return hls2hex(hue, lightness, DEFAULT_COLOR_SATURATION)
heatmap_data = {}
max_value, min_value = None, None
for line in open(args.heatmap):
if line.startswith('#COLNAMES'):
pass
elif line.startswith('#') or not line.strip():
pass
else:
fields = line.split('\t')
name = fields[0].strip()
values = [float(x) if x else None for x in fields[1:]]
maxv = max(values)
minv = min(values)
if max_value is None or maxv > max_value:
max_value = maxv
if min_value is None or minv < min_value:
min_value = minv
heatmap_data[name] = values
heatmap_center_value = 0
heatmap_color_center = "white"
heatmap_color_up = 0.3
heatmap_color_down = 0.7
heatmap_color_missing = "black"
heatmap_max_value = abs(heatmap_center_value - max_value)
heatmap_min_value = abs(heatmap_center_value - min_value)
if heatmap_center_value <= min_value:
heatmap_max_value = heatmap_min_value + heatmap_max_value
else:
heatmap_max_value = max(heatmap_min_value, heatmap_max_value)
# scale the tree
if not args.height:
args.height = None
if not args.width:
args.width = None
f2color = {}
f2last_seed = {}
for node in t.traverse():
node.img_style['size'] = 0
if len(node.children) == 1:
node.img_style['size'] = 2
node.img_style['shape'] = "square"
node.img_style['fgcolor'] = "steelblue"
ftype_pos = defaultdict(int)
for findex, f in enumerate(FACES):
if (f['nodetype'] == 'any' or
(f['nodetype'] == 'leaf' and node.is_leaf()) or
(f['nodetype'] == 'internal' and not node.is_leaf())):
# if node passes face filters
if node_matcher(node, f["filters"]):
if f["value"].startswith("@"):
fvalue = getattr(node, f["value"][1:], None)
else:
fvalue = f["value"]
# if node's attribute has content, generate face
if fvalue is not None:
fsize = f["size"]
fbgcolor = f["bgcolor"]
fcolor = f['color']
if fcolor:
# Parse color options
auto_m = re.search("auto\(([^)]*)\)", fcolor)
if auto_m:
target_attr = auto_m.groups()[0].strip()
if not target_attr :
color_keyattr = f["value"]
else:
color_keyattr = target_attr
color_keyattr = color_keyattr.lstrip('@')
color_bin = getattr(node, color_keyattr, None)
last_seed = f2last_seed.setdefault(color_keyattr, random.random())
seed = last_seed + 0.10 + random.uniform(0.1, 0.2)
f2last_seed[color_keyattr] = seed
fcolor = f2color.setdefault(color_bin, random_color(h=seed))
if fbgcolor:
# Parse color options
auto_m = re.search("auto\(([^)]*)\)", fbgcolor)
if auto_m:
target_attr = auto_m.groups()[0].strip()
if not target_attr :
color_keyattr = f["value"]
else:
color_keyattr = target_attr
color_keyattr = color_keyattr.lstrip('@')
color_bin = getattr(node, color_keyattr, None)
last_seed = f2last_seed.setdefault(color_keyattr, random.random())
seed = last_seed + 0.10 + random.uniform(0.1, 0.2)
f2last_seed[color_keyattr] = seed
fbgcolor = f2color.setdefault(color_bin, random_color(h=seed))
if f["ftype"] == "text":
if f.get("format", None):
fvalue = f["format"] % fvalue
F = TextFace(fvalue,
fsize = fsize,
fgcolor = fcolor or "black",
fstyle = f.get('fstyle', None))
elif f["ftype"] == "fullseq":
F = faces.SeqMotifFace(seq=fvalue, seq_format="seq",
seqtail_format="seq",
height=fsize)
elif f["ftype"] == "compactseq":
F = faces.SeqMotifFace(seq=fvalue, seq_format="compactseq",
seqtail_format="compactseq",
height=fsize)
elif f["ftype"] == "blockseq":
F = faces.SeqMotifFace(seq=fvalue, seq_format="blockseq",
seqtail_format="blockseq",
height=fsize,
fgcolor=fcolor or "slategrey",
bgcolor=fbgcolor or "slategrey",
scale_factor = 1.0)
fbgcolor = None
elif f["ftype"] == "bubble":
try:
v = float(fvalue)
except ValueError:
rad = fsize
else:
rad = fsize * v
F = faces.CircleFace(radius=rad, style="sphere",
color=fcolor or "steelblue")
elif f["ftype"] == "heatmap":
if not f['column']:
col = ftype_pos[f["pos"]]
else:
col = f["column"]
for i, value in enumerate(heatmap_data.get(node.name, [])):
ftype_pos[f["pos"]] += 1
if value is None:
color = heatmap_color_missing
elif value > heatmap_center_value:
color = gradient_color(abs(heatmap_center_value - value), heatmap_max_value, hue=heatmap_color_up)
elif value < heatmap_center_value:
color = gradient_color(abs(heatmap_center_value - value), heatmap_max_value, hue=heatmap_color_down)
else:
color = heatmap_color_center
node.add_face(RectFace(20, 20, color, color), position="aligned", column=col + i)
# Add header
# for i, name in enumerate(header):
# nameF = TextFace(name, fsize=7)
# nameF.rotation = -90
# tree_style.aligned_header.add_face(nameF, column=i)
F = None
elif f["ftype"] == "profile":
# internal profiles?
F = None
elif f["ftype"] == "barchart":
F = None
elif f["ftype"] == "piechart":
F = None
# Add the Face
if F:
F.opacity = f['opacity'] or 1.0
# Set face general attributes
if fbgcolor:
F.background.color = fbgcolor
if not f['column']:
col = ftype_pos[f["pos"]]
ftype_pos[f["pos"]] += 1
else:
col = f["column"]
node.add_face(F, column=col, position=f["pos"])
if args.image:
t.render("t%d.%s" %(tindex, args.image),
tree_style=ts, w=args.width, h=args.height, units=args.size_units)
else:
t.show(None, tree_style=ts)
