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 styleFace(val):
x = TextFace(val)
x.margin_bottom = 5
x.margin_right = 10
x.rotation = 270
x.fsize = 6
return x
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 format_nodes(node, node_style, sus_clades, t):
"""This function visually formats the nodes in the svg file based on the nodes' support values and whether or not
the clade is suspicous"""
supp = TextFace(f'{int(node.support)}', fsize=8)
if node.support >= 70:
supp.bold = True
taxons = set()
orgs = node.get_leaf_names()
if len(orgs) > 1:
for org in orgs:
if '..' in org: # for paralogs
org = org.split('..')[0]
else:
org = org.split('_')[0] # for potential orthologs
taxons.add(metadata[org]['Higher Taxonomy'])
if len(taxons) > 1 and (len(node) < (len(t) / 2)):
node_style['shape'] = 'sphere'
node_style['size'] = 12
node_style['fgcolor'] = 'red'
node_style['bgcolor'] = 'Silver'
sus_clades += 1
else:
supp.fsize = 7
return supp, sus_clades
time_end = trackResult[r, 2]
parent_id = trackResult[r, 3]
time_duration = np.abs(time_begin-time_end)
# for root
if parent_id == 0:
# Add name to root for the first iteration
root.add_feature("name", str(cell_id))
# change the branch length
root.add_feature("dist", time_duration)
#change node style
root.set_style(ns_root)
# set node name to face
nameFace = TextFace(root.name)
nameFace.fgcolor = "white"
nameFace.fsize = 15
# nameFace.border.width = 1
nameFace.background.color = "green"
node_cur.add_face(nameFace, column=1, position="branch-bottom")
else: # for child
#### search the parent node by parent_id
node_cur = root.search_nodes(name=str(parent_id))
# there should be only one parent node
if len(node_cur) == 1:
#### set child with its id
node_cur = node_cur[0].add_child(name=str(cell_id))
#### set duration
node_cur.add_feature("dist", time_duration)
# set node style
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
time_end = trackResult[r, 2]
parent_id = trackResult[r, 3]
time_duration = np.abs(time_begin - time_end)
# for root
if parent_id == 0:
# Add name to root for the first iteration
root.add_feature("name", str(cell_id))
# change the branch length
root.add_feature("dist", time_duration)
#change node style
root.set_style(ns_root)
# set node name to face
nameFace = TextFace(root.name)
nameFace.fgcolor = "white"
nameFace.fsize = 15
# nameFace.border.width = 1
nameFace.background.color = "green"
node_cur.add_face(nameFace, column=1, position="branch-bottom")
else: # for child
#### search the parent node by parent_id
node_cur = root.search_nodes(name=str(parent_id))
# there should be only one parent node
if len(node_cur) == 1:
#### set child with its id
node_cur = node_cur[0].add_child(name=str(cell_id))
#### set duration
node_cur.add_feature("dist", time_duration)
# set node style
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