def draw(self,
file,
colors,
color_internal_nodes=True,
legend_labels=(),
show_branch_support=True,
show_scale=True,
legend_scale=1,
mode="c",
neighbours=None,
neighbours_block=None):
max_color = len(colors)
used_colors = set()
for node in self.tree.traverse():
if not (color_internal_nodes or node.is_leaf()): continue
color = colors[min(node.color, max_color - 1)]
node.img_style['bgcolor'] = color
used_colors.add(color)
ts = TreeStyle()
ts.mode = mode
ts.scale = self.scale
# Disable the default tip names config
ts.show_leaf_name = False
ts.show_branch_support = show_branch_support
# ts.branch_vertical_margin = 20
ts.show_scale = show_scale
cur_max_color = max(v.color for v in self.tree.traverse())
current_colors = colors[0:cur_max_color + 1]
for i, (label, color_) in enumerate(zip(legend_labels,
current_colors)):
if color_ not in used_colors: continue
rf = RectFace(20 * legend_scale, 16 * legend_scale, color_, color_)
rf.inner_border.width = 1
rf.margin_right = 14
rf.margin_left = 14
tf = TextFace(label, fsize=26 * legend_scale)
tf.margin_right = 14
ts.legend.add_face(rf, column=0)
ts.legend.add_face(tf, column=1)
if neighbours:
old_tree = self.tree.copy()
self.draw_neighbours(neighbours, neighbours_block)
self.tree.render(file, w=1000, tree_style=ts)
if neighbours:
self.tree = old_tree
def get_tree_style():
ts = TreeStyle()
ts.show_leaf_name = False # True
ts.layout_fn = layout
ts.rotation = 90
ts.branch_vertical_margin = 10
ts.show_scale = False
# ts.mode = "c"
ts.scale = 50
ts.title.add_face(TextFace(" ", fsize=20), column=0)
# for n in t.traverse():
# nstyle = NodeStyle()
# nstyle["fgcolor"] = "red"
# nstyle["size"] = 15
# n.set_style(nstyle)
# ts.show_leaf_name = True
ts.legend.add_face(TextFace(" "), column=0)
ts.legend.add_face(TextFace(" "), column=1)
ts.legend.add_face(RectFace(20, 20, NO_SUPPORT_COLOR, NO_SUPPORT_COLOR),
column=0)
ts.legend.add_face(TextFace(" Topic with no support (u(t)=0)"), column=1)
ts.legend.add_face(TextFace(" "), column=0)
ts.legend.add_face(TextFace(" "), column=1)
ts.legend.add_face(RectFace(20, 20, "#90ee90", "#90ee90"), column=0)
ts.legend.add_face(TextFace(" Topic with minor support 0<u(t)<=0.4"),
column=1)
ts.legend.add_face(TextFace(" "), column=0)
ts.legend.add_face(TextFace(" "), column=1)
ts.legend.add_face(RectFace(20, 20, "green", "green"), column=0)
ts.legend.add_face(
TextFace(u" Topic with medium support 0.4<u(t)<=0.6 "), column=1)
ts.legend.add_face(TextFace(" "), column=0)
ts.legend.add_face(TextFace(" "), column=1)
ts.legend.add_face(RectFace(20, 20, "#004000", "#004000"), column=0)
ts.legend.add_face(TextFace(" Topic with high support u(t)>0.6"),
column=1)
ts.legend.add_face(TextFace(" "), column=0)
ts.legend.add_face(TextFace(" "), column=1)
ts.legend.add_face(CircleFace(10, "red"), column=0)
ts.legend.add_face(TextFace(" Gap"), column=1)
ts.legend.add_face(TextFace(" "), column=0)
ts.legend.add_face(TextFace(" "), column=1)
ts.legend.add_face(RectFace(40, 40, "#004000", "#004000"),
column=0) # green
# ts.legend.add_face(CircleFace(15, "green"), column=1)
ts.legend.add_face(TextFace(" Head subject or offshoot"), column=1)
ts.legend_position = 4
# ts.title.add_face(TextFace(" ", fsize=20), column=0)
return ts
def render_tree(tree, mode, cluster, colourDict, width, position):
duplicates = ["'EBOV|EMLab-RT|KG12||GIN|Boke|?|MinION_LQ05|2015-05-27'", "'EBOV|EMLab-RT|KG45||GIN|Boke|?|MinION_LQ10|2015-06-09'", "'EBOV|EMLab-RT|KG90||GIN|Boke|?|MinION_LQ05|2015-06-19'", "'EBOV|EMLab-RT|KG91||GIN|Boke|?|MinION_LQ05|2015-06-20'"]
print('Running %s: %s cluster' %(mode, cluster))
if mode == 'small':
#delete unwanted leaves
keep_leaves = []
b = ["'" + clusters[cluster][0] + "'", "'" + clusters[cluster][1] + "'"]
for a in tree.get_common_ancestor(b).get_leaves():
keep_leaves.append(a.name)
delete_leaves = [leaf for leaf in tree.get_leaf_names() if leaf not in keep_leaves]
#if cluster == 'Boke':
# delete_leaves.extend(duplicates)
print('Keeping %s leaves' %len(keep_leaves))
for leaf in delete_leaves:
if tree.search_nodes(name=leaf)[0]:
n = tree.search_nodes(name=leaf)[0]
#if leaf in duplicates:
# n.delete(preserve_branch_length=True)
#else:
n.delete()
#print 'Removed %s' %(n.get_leaf_names()[0])
tree.ladderize()
tree.write(outfile=sys.argv[1] + '_' + cluster + '.nwk')
for n in tree.get_leaves():
display_name = n.get_leaf_names()[0][1:-1]
country = metadata[display_name]['country']
instrument = metadata[display_name]['instrument']
#if cluster == 'Boke':
if mode == 'big':
n.add_face(MyTextFace(metadata[display_name]['short_id'], ftype="Helvetica", fsize=size[mode]), column=0, position="aligned")
n.add_face(MyTextFace(metadata[display_name]['prefec'], ftype="Helvetica", fsize=size[mode]), column=1, position="aligned")
n.add_face(MyTextFace(metadata[display_name]['date'], ftype="Helvetica", fsize=size[mode]), column=2, position="aligned")
if country == 'GUI' or country == 'GIN':
if instrument == 'MinION':
C = CircleFace(radius=size[mode]/2, color=colourDict[metadata[display_name]['prefec']])
else:
C = CircleFace(radius=size[mode]/2, color='#F1F1F1')
elif country == 'SLE':
if instrument == 'MinION':
C = RectFace(width=size[mode], height=size[mode], bgcolor=colourDict[metadata[display_name]['prefec']], fgcolor='#FFFFFF')
else:
C = RectFace(width=size[mode], height=size[mode], bgcolor='#D3D3D3', fgcolor='#FFFFFF')
elif country == 'LIB' or country == 'LBR':
if instrument == 'MinION':
C = RectFace(triangle=True, width=size[mode], height=size[mode], bgcolor=colourDict[metadata[display_name]['prefec']], fgcolor='#FFFFFF')
else:
C = RectFace(triangle=True, width=size[mode], height=size[mode], bgcolor='#939393', fgcolor='#FFFFFF')
n.add_face(C, column=1, position=position)
tree.render(file_name=sys.argv[1] + '_' + cluster + '.pdf', tree_style=ts, w=width)
def layout(node):
if node.is_leaf():
N = AttrFace("sci_name", fsize=30)
faces.add_face_to_node(N, node, 0, position="branch-right")
circle = RectFace(100, 100, "black", "black")
faces.add_face_to_node(circle, position="aligned", column=0, node=node)
if "coli" in node.sci_name:
circle = RectFace(100, 100, "red", "red")
faces.add_face_to_node(circle,
position="aligned",
column=1,
node=node)
if "Salm" in node.sci_name:
circle = RectFace(100, 100, "blue", "blue")
faces.add_face_to_node(circle,
position="aligned",
column=2,
node=node)
def get_example_tree():
# Performs a tree reconciliation analysis
gene_tree_nw = '((Dme_001,Dme_002),(((Cfa_001,Mms_001),((Hsa_001,Ptr_001),Mmu_001)),(Ptr_002,(Hsa_002,Mmu_002))));'
t = PhyloTree(gene_tree_nw)
ts = TreeStyle()
# disable default PhyloTree Layout
ts.layout_fn = lambda x: True
t.link_to_alignment(alg)
node2content = t.get_cached_content()
for node in t.traverse():
node.img_style["size"] = 0
if not node.is_leaf():
leaves = node2content[node]
# get columns with different aa
subseqs, relevant_columns = mutation_columns([lf.sequence for lf in leaves])
for seq in subseqs:
f = SeqMotifFace(seq, seq_format="seq", width=10, height=8)
f.margin_top = 2
f.margin_right = 6
node.add_face(f, column=0, position="branch-bottom")
for j, col in enumerate(relevant_columns):
col_f = RectFace(10, 10, fgcolor=None, bgcolor=None,
label={"text":str(col), "fonttype":"Courier", "color":"black", "fontsize":6})
node.add_face(col_f, column=j, position="branch-top")
col_f.margin_bottom = 2
else:
f = SeqMotifFace(node.sequence, seq_format="seq", width=6)
node.add_face(f, column=0, position="aligned")
alg_length = len(lf.sequence)
ts.draw_aligned_faces_as_table = False
for colnum in xrange(alg_length):
col_f = RectFace(10, 10, fgcolor=None, bgcolor=None,
label={"text":str(colnum), "fonttype":"Courier", "color":"black", "fontsize":6})
ts.aligned_header.add_face(col_f, column=colnum)
return t, ts
def layout(node):
if node.is_leaf():
if node.name=="A":
node.color="Blue"
node.attr2="Green"
elif node.name=="B":
node.color="Red"
node.attr2="Orange"
else:
node.color="Grey"
node.attr2="Yellow"
faces.add_face_to_node( CircleFace(radius=5, color=node.color), node, column=1)
faces.add_face_to_node( RectFace(width=10, height=10, fgcolor=node.attr2, bgcolor=node.attr2), node, column=2)
def draw_tree(pstrains, names=False, subset=None):
evol = {x.rstrip() for x in open('input/evolution_experiment.txt')}
tree = get_tree('input/tree.nwk')
strains = {x.name for x in tree.traverse() if x.name != ''}
if subset is None:
subset = strains
evol_tree = {x.name for x in tree.traverse() if x.name in evol}
commensals = {}
for x in strains:
if pstrains[x] == 'Commensal strain':
commensals[x] = colors.cnames['blue']
elif pstrains[x] == 'Pathogenic strain':
commensals[x] = colors.cnames['red']
else:
commensals[x] = colors.cnames['white']
ref = NodeStyle()
ref['fgcolor'] = sns.xkcd_rgb['light red']
inner = NodeStyle()
inner['size'] = 0
for n in tree.traverse():
if not n.is_leaf():
n.set_style(inner)
continue
if n.name not in subset:
continue
if not names:
r = RectFace(10, 3, commensals[n.name], commensals[n.name])
n.add_face(r, 0, position="aligned")
ev = NodeStyle()
ev["fgcolor"] = "black"
ev['size'] = 3
n.set_style(ev)
circular_style = TreeStyle()
circular_style.mode = "c"
if not names:
circular_style.scale = 1300
else:
circular_style.scale = 7300
circular_style.show_leaf_name = names
return tree, circular_style
def visualize_tree(ete_tree_obj,
genotypes={},
metadata={},
cansnp_supported_nodes={}):
"""
Function for visualization of cannonical snp supported clades
Parameters
----------
ete_tree_obj
cansnp_supported_nodes
Returns
-------
"""
# Basic tree style
ts = TreeStyle()
ts.show_leaf_name = True
# Draws nodes as small red spheres of diameter equal to 10 pixels
for n in ete_tree_obj.traverse():
nstyle = NodeStyle()
nstyle["shape"] = "sphere"
nstyle["fgcolor"] = "darkred"
if n.name in cansnp_supported_nodes:
print("Found")
#nstyle["fgcolor"] = "blue"
#nstyle["size"] = cansnp_supported_nodes[n.name]['size']
#nstyle["bgcolor"] =cansnp_supported_nodes[n.name]['bgcolor']
face = RectFace(width=100,
height=100,
fgcolor='red',
bgcolor='green')
n.add_face(face, column=1)
n.set_style(nstyle)
ete_tree_obj.show(tree_style=ts)
def setup_heatmap(tree,
tree_style,
header,
center_value=0,
nameMap={},
nameLabel='',
color_up=0.7,
color_down=0.2,
color_center="white"):
DEFAULT_COLOR_SATURATION = 0.5
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(map(lambda x: int(x * 255), colorsys.hls_to_rgb(h, l,
s))))
lightness = 1 - (value * BASE_LIGHTNESS) / max_value
return hls2hex(hue, lightness, DEFAULT_COLOR_SATURATION)
# Calculate max gradient value from the ClusterTree matrix
maxv = abs(center_value - tree.arraytable._matrix_max)
minv = abs(center_value - tree.arraytable._matrix_min)
if center_value <= tree.arraytable._matrix_min:
MAX_VALUE = minv + maxv
else:
MAX_VALUE = max(maxv, minv)
# Add heatmap colors to tree
cols_add_before_heat = 0
if nameMap:
cols_add_before_heat = 1
for lf in tree:
if nameMap:
longNameFace = faces.TextFace(nameMap.get(lf.name, lf.name))
lf.add_face(longNameFace, column=0, position="aligned")
for i, value in enumerate(getattr(lf, "profile", [])):
if value > center_value:
color = gradient_color(abs(center_value - value),
MAX_VALUE,
hue=color_up)
elif value < center_value:
color = gradient_color(abs(center_value - value),
MAX_VALUE,
hue=color_down)
else:
color = center_value
#color = "LightGreen"
lf.add_face(RectFace(25, 25, "white", color),
position="aligned",
column=i + cols_add_before_heat)
# Uncomment to add numeric values to the matrix
#lf.add_face(TextFace("%0.2f "%value, fsize=5), position="aligned", column=i)
lf.add_face(nameFace,
column=i + cols_add_before_heat + 1,
position="aligned")
if nameMap and nameLabel:
nameF = TextFace(nameLabel, fsize=10)
#nameF.rotation = -90
tree_style.aligned_header.add_face(nameF, column=0)
# Add header
for i, name in enumerate(header):
nameF = TextFace(name, fsize=10)
nameF.rotation = -90
tree_style.aligned_header.add_face(nameF,
column=i + cols_add_before_heat)
def draw_tree(tree, conf, outfile):
try:
from ete3 import (add_face_to_node, AttrFace, TextFace, TreeStyle,
RectFace, CircleFace, SequenceFace, random_color,
SeqMotifFace)
except ImportError as e:
print(e)
return
def ly_basic(node):
if node.is_leaf():
node.img_style['size'] = 0
else:
node.img_style['size'] = 0
node.img_style['shape'] = 'square'
if len(MIXED_RES) > 1 and hasattr(node, "tree_seqtype"):
if node.tree_seqtype == "nt":
node.img_style["bgcolor"] = "#CFE6CA"
ntF = TextFace("nt",
fsize=6,
fgcolor='#444',
ftype='Helvetica')
add_face_to_node(ntF, node, 10, position="branch-bottom")
if len(NPR_TREES) > 1 and hasattr(node, "tree_type"):
node.img_style['size'] = 4
node.img_style['fgcolor'] = "steelblue"
node.img_style['hz_line_width'] = 1
node.img_style['vt_line_width'] = 1
def ly_leaf_names(node):
if node.is_leaf():
spF = TextFace(node.species,
fsize=10,
fgcolor='#444444',
fstyle='italic',
ftype='Helvetica')
add_face_to_node(spF, node, column=0, position='branch-right')
if hasattr(node, 'genename'):
geneF = TextFace(" (%s)" % node.genename,
fsize=8,
fgcolor='#777777',
ftype='Helvetica')
add_face_to_node(geneF,
node,
column=1,
position='branch-right')
def ly_supports(node):
if not node.is_leaf() and node.up:
supFace = TextFace("%0.2g" % (node.support),
fsize=7,
fgcolor='indianred')
add_face_to_node(supFace, node, column=0, position='branch-top')
def ly_tax_labels(node):
if node.is_leaf():
c = LABEL_START_COL
largest = 0
for tname in TRACKED_CLADES:
if hasattr(node,
"named_lineage") and tname in node.named_lineage:
linF = TextFace(tname, fsize=10, fgcolor='white')
linF.margin_left = 3
linF.margin_right = 2
linF.background.color = lin2color[tname]
add_face_to_node(linF, node, c, position='aligned')
c += 1
for n in range(c, len(TRACKED_CLADES)):
add_face_to_node(TextFace('', fsize=10, fgcolor='slategrey'),
node,
c,
position='aligned')
c += 1
def ly_full_alg(node):
pass
def ly_block_alg(node):
if node.is_leaf():
if 'sequence' in node.features:
seqFace = SeqMotifFace(node.sequence, [])
# [10, 100, "[]", None, 10, "black", "rgradient:blue", "arial|8|white|domain Name"],
motifs = []
last_lt = None
for c, lt in enumerate(node.sequence):
if lt != '-':
if last_lt is None:
last_lt = c
if c + 1 == len(node.sequence):
start, end = last_lt, c
motifs.append([
start, end, "()", 0, 12, "slategrey",
"slategrey", None
])
last_lt = None
elif lt == '-':
if last_lt is not None:
start, end = last_lt, c - 1
motifs.append([
start, end, "()", 0, 12, "grey", "slategrey",
None
])
last_lt = None
seqFace = SeqMotifFace(node.sequence,
motifs,
intermotif_format="line",
seqtail_format="line",
scale_factor=ALG_SCALE)
add_face_to_node(seqFace, node, ALG_START_COL, aligned=True)
TRACKED_CLADES = [
"Eukaryota",
"Viridiplantae",
"Fungi",
"Alveolata",
"Metazoa",
"Stramenopiles",
"Rhodophyta",
"Amoebozoa",
"Crypthophyta",
"Bacteria",
"Alphaproteobacteria",
"Betaproteobacteria",
"Cyanobacteria",
"Gammaproteobacteria",
]
# ["Opisthokonta", "Apicomplexa"]
colors = random_color(num=len(TRACKED_CLADES), s=0.45)
lin2color = dict([(ln, colors[i]) for i, ln in enumerate(TRACKED_CLADES)])
NAME_FACE = AttrFace('name', fsize=10, fgcolor='#444444')
LABEL_START_COL = 10
ALG_START_COL = 40
ts = TreeStyle()
ts.draw_aligned_faces_as_table = False
ts.draw_guiding_lines = False
ts.show_leaf_name = False
ts.show_branch_support = False
ts.scale = 160
ts.layout_fn = [ly_basic, ly_leaf_names, ly_supports, ly_tax_labels]
MIXED_RES = set()
MAX_SEQ_LEN = 0
NPR_TREES = []
for n in tree.traverse():
if hasattr(n, "tree_seqtype"):
MIXED_RES.add(n.tree_seqtype)
if hasattr(n, "tree_type"):
NPR_TREES.append(n.tree_type)
seq = getattr(n, "sequence", "")
MAX_SEQ_LEN = max(len(seq), MAX_SEQ_LEN)
if MAX_SEQ_LEN:
ALG_SCALE = min(1, 1000. / MAX_SEQ_LEN)
ts.layout_fn.append(ly_block_alg)
if len(NPR_TREES) > 1:
rF = RectFace(4, 4, "steelblue", "steelblue")
rF.margin_right = 10
rF.margin_left = 10
ts.legend.add_face(rF, 0)
ts.legend.add_face(TextFace(" NPR node"), 1)
ts.legend_position = 3
if len(MIXED_RES) > 1:
rF = RectFace(20, 20, "#CFE6CA", "#CFE6CA")
rF.margin_right = 10
rF.margin_left = 10
ts.legend.add_face(rF, 0)
ts.legend.add_face(TextFace(" Nucleotide based alignment"), 1)
ts.legend_position = 3
try:
tree.set_species_naming_function(spname)
annotate_tree_with_ncbi(tree)
a = tree.search_nodes(species='Dictyostelium discoideum')[0]
b = tree.search_nodes(species='Chondrus crispus')[0]
#out = tree.get_common_ancestor([a, b])
#out = tree.search_nodes(species='Haemophilus parahaemolyticus')[0].up
tree.set_outgroup(out)
tree.swap_children()
except Exception:
pass
tree.render(outfile, tree_style=ts, w=170, units='mm', dpi=150)
tree.render(outfile + '.svg', tree_style=ts, w=170, units='mm', dpi=150)
tree.render(outfile + '.pdf', tree_style=ts, w=170, units='mm', dpi=150)
def plot(self, placement, togjson, outdir, cfg):
"""
plot a plcement in the tree
show all pplacer placements and the LCA and HCA node
as well as the inferred lineage
"""
from ete3 import NodeStyle, TreeStyle
from ete3 import CircleFace, TextFace, RectFace
logging.debug("Plotting trees now")
# with no X display this needs to be set
os.environ["QT_QPA_PLATFORM"] = "offscreen"
info = self.loadInfo(togjson)
def defaultNodeStyle():
return NodeStyle()
nodeStyles = defaultdict(defaultNodeStyle)
no = 0
for LCAp, HPAp in zip(placement["LCA"], placement["HPA"]):
plotpath = os.path.join(outdir, f"tree_{no}.png")
# make shallow copy
t = self.t
LCA = LCAp["node"]
HPA = HPAp["node"]
# define basic tree style
ts = TreeStyle()
# hide leave names
ts.show_leaf_name = False
ts.root_opening_factor = 1
# circular tree
ts.mode = "c"
ts.rotation = 210
ts.arc_start = 0 # 0 degrees = 3 o'clock
ts.arc_span = 350
highlightsize = 80
nodesize = 10
# define styles for special nodes
# at the moment hard coded, but could be accesible for the user
# LCA style
LCAstyle = NodeStyle()
LCAstyle["fgcolor"] = "#33a02c"
LCAstyle["bgcolor"] = "#b2df8a"
LCAstyle["size"] = highlightsize
# HPA style
HPAstyle = NodeStyle()
HPAstyle["fgcolor"] = "#1f78b4"
HPAstyle["bgcolor"] = "#a6cee3"
HPAstyle["size"] = highlightsize
# default node
defaultStyle = NodeStyle()
defaultStyle["fgcolor"] = "gray"
defaultStyle["size"] = nodesize
# add legend
ts.legend_position = 1
ts.legend.add_face(CircleFace(40, LCAstyle["fgcolor"]), column=1)
ts.legend.add_face(TextFace(f"LCA", fsize=50), column=2)
ts.legend.add_face(CircleFace(40, HPAstyle["fgcolor"]), column=1)
ts.legend.add_face(TextFace(f"HPA", fsize=50), column=2)
i = 1
ts.legend.add_face(TextFace(f"p = {i}", fsize=50), column=1)
while i > 0:
temp_face = RectFace(60,
10,
fgcolor=p_to_color(i),
bgcolor=p_to_color(i))
temp_face.margin_top = -4
ts.legend.add_face(temp_face, column=1)
i -= 0.01
ts.legend.add_face(TextFace(f"p = {cfg['minPlacementLikelyhood']}",
fsize=50),
column=1)
# add highlights for each placed protein
for n in t.traverse():
if n.name.startswith("PTHR"):
# set color based on posterior prob:
x = (info[n.name]["post_prob"] -
cfg["minPlacementLikelyhood"]) / (
1 - cfg["minPlacementLikelyhood"])
# orange to purple gradient from 0 to 1 posterior propability
he = p_to_color(x)
nodeStyles[he]["bgcolor"] = he
# define back color of locations
n.set_style(nodeStyles[he])
elif n.name == LCA:
n.set_style(LCAstyle)
elif n.name == HPA:
n.set_style(HPAstyle)
else:
n.set_style(defaultStyle)
# plot to disk
_ = t.render(plotpath, w=320, units="mm", tree_style=ts)
no = no + 1
def heatmap_view(tree, orthologous_groups, save_dir):
"""Generates a heatmap of regulation states in all species."""
light_tree = copy.deepcopy(tree) # Tree copy for the light heatmap
# Heat map settings
rect_face_fgcolor = 'black'
locus_tag_len = max(
len(gene.locus_tag) + 5 for ortho_grp in orthologous_groups
for gene in ortho_grp.genes)
rect_face_width = locus_tag_len * 8
light_rect_face_width = 20
rect_face_height = 20
rotation = 90
# Sort orthologous groups by the number of regulated genes in each group
orthologous_groups = filter_and_sort_orthologous_grps(orthologous_groups)
# For each species and its gene in each orthologous group, draw a rectangle
for node, light_node in zip(tree.get_leaves(), light_tree.get_leaves()):
for i, orthologous_grp in enumerate(orthologous_groups, start=1):
#get all orthologs in group
matching_genes = [g for g in orthologous_grp.genes \
if g.genome.strain_name == node.name]
#if there is ortholog
if len(matching_genes) > 0:
# Get the first ortholog from the genome in the group
#this is the one with higher probability of regulation.
#so this probability will be displayed for the group
gene = matching_genes[0]
p_regulation = gene.operon.regulation_probability
p_notregulation = 1.0 - p_regulation
p_absence = 0
# No ortholog from this genome
else:
gene = None
p_regulation = 0
p_notregulation = 0
p_absence = 1
# Color of the rectangle is based on probabilities
rect_face_bgcolor = rgb2hex(p_notregulation, p_regulation,
p_absence)
rect_face_text = ('%s [%d]' %
(gene.locus_tag, gene.operon.operon_id)
if gene else '')
rect_face_label = {
'text': rect_face_text,
'font': 'Courier',
'fontsize': 8,
'color': 'black'
}
# Create the rectangle
rect_face = RectFace(rect_face_width,
rect_face_height,
rect_face_fgcolor,
rect_face_bgcolor,
label=rect_face_label)
light_rect_face = RectFace(light_rect_face_width,
rect_face_height,
rect_face_fgcolor,
rect_face_bgcolor,
label='')
rect_face.rotation = -rotation
light_rect_face.rotation = -rotation
# Add the rectangle to the corresponding column
node.add_face(rect_face, column=i, position='aligned')
light_node.add_face(light_rect_face, column=i, position='aligned')
ts = TreeStyle()
# Add orthologous group descriptions
descriptions = ['-'.join([grp.description, \
str([item['ID'] for item in grp.COGs]) if len(grp.COGs)>0 else '', \
str([item['ID'] for item in grp.NOGs]) if len(grp.NOGs)>0 else '', \
str([item['ID'] for item in grp.PFAMs])] if len(grp.PFAMs)>0 else '')\
for grp in orthologous_groups]
max_description_len = max(map(len, descriptions))
descriptions = [
'[%d]' % i + description + ' ' *
(max_description_len - len(description))
for i, description in enumerate(descriptions, start=1)
]
for i, description in enumerate(descriptions, start=1):
text_face = TextFace(description, ftype='Courier')
text_face.hz_align = 1
text_face.vt_align = 1
text_face.rotation = -rotation
ts.aligned_header.add_face(text_face, column=i)
# Rotate the generated heatmap.
ts.margin_left = 10
ts.margin_top = 20
ts.rotation = rotation
ts.show_scale = False
# For some reason, it can't render to PDF in color
tree.render(os.path.join(save_dir, 'heatmap.svg'), tree_style=ts)
light_tree.render(os.path.join(save_dir, 'heatmap_light.svg'),
tree_style=ts)
def draw_tree(the_tree, colour, back_color, label, out_file, the_scale, extend,
bootstrap, group_file, grid_options, the_table, pres_abs,
circular):
t = Tree(the_tree, quoted_node_names=True)
# t.ladderize()
font_size = 8
font_type = 'Heveltica'
font_gap = 3
font_buffer = 10
o = t.get_midpoint_outgroup()
t.set_outgroup(o)
the_leaves = []
for leaves in t.iter_leaves():
the_leaves.append(leaves)
groups = {}
num = 0
# set cutoff value for clades as 1/20th of the distance between the furthest two branches
# assign nodes to groups
last_node = None
ca_list = []
if not group_file is None:
style = NodeStyle()
style['size'] = 0
style["vt_line_color"] = '#000000'
style["hz_line_color"] = '#000000'
style["vt_line_width"] = 1
style["hz_line_width"] = 1
for n in t.traverse():
n.set_style(style)
with open(group_file) as f:
group_dict = {}
for line in f:
group_dict[line.split()[0]] = line.split()[1]
for node in the_leaves:
i = node.name
for j in group_dict:
if j in i:
if group_dict[j] in groups:
groups[group_dict[j]].append(i)
else:
groups[group_dict[j]] = [i]
coloured_nodes = []
for i in groups:
the_col = i
style = NodeStyle()
style['size'] = 0
style["vt_line_color"] = the_col
style["hz_line_color"] = the_col
style["vt_line_width"] = 2
style["hz_line_width"] = 2
if len(groups[i]) == 1:
ca = t.search_nodes(name=groups[i][0])[0]
ca.set_style(style)
coloured_nodes.append(ca)
else:
ca = t.get_common_ancestor(groups[i])
ca.set_style(style)
coloured_nodes.append(ca)
tocolor = []
for j in ca.children:
tocolor.append(j)
while len(tocolor) > 0:
x = tocolor.pop(0)
coloured_nodes.append(x)
x.set_style(style)
for j in x.children:
tocolor.append(j)
ca_list.append((ca, the_col))
if back_color:
# for each common ancestor node get it's closest common ancestor neighbour and find the common ancestor of those two nodes
# colour the common ancestor then add it to the group - continue until only the root node is left
while len(ca_list) > 1:
distance = float('inf')
for i, col1 in ca_list:
for j, col2 in ca_list:
if not i is j:
parent = t.get_common_ancestor(i, j)
getit = True
the_dist = t.get_distance(i, j)
if the_dist <= distance:
distance = the_dist
the_i = i
the_j = j
the_i_col = col1
the_j_col = col2
ca_list.remove((the_i, the_i_col))
ca_list.remove((the_j, the_j_col))
rgb1 = strtorgb(the_i_col)
rgb2 = strtorgb(the_j_col)
rgb3 = ((rgb1[0] + rgb2[0]) / 2, (rgb1[1] + rgb2[1]) / 2,
(rgb1[2] + rgb2[2]) / 2)
new_col = colorstr(rgb3)
new_node = t.get_common_ancestor(the_i, the_j)
the_col = new_col
style = NodeStyle()
style['size'] = 0
style["vt_line_color"] = the_col
style["hz_line_color"] = the_col
style["vt_line_width"] = 2
style["hz_line_width"] = 2
new_node.set_style(style)
coloured_nodes.append(new_node)
ca_list.append((new_node, new_col))
for j in new_node.children:
tocolor.append(j)
while len(tocolor) > 0:
x = tocolor.pop(0)
if not x in coloured_nodes:
coloured_nodes.append(x)
x.set_style(style)
for j in x.children:
tocolor.append(j)
elif colour:
distances = []
for node1 in the_leaves:
for node2 in the_leaves:
if node1 != node2:
distances.append(t.get_distance(node1, node2))
distances.sort()
clade_cutoff = distances[len(distances) / 4]
for node in the_leaves:
i = node.name
if not last_node is None:
if t.get_distance(node, last_node) <= clade_cutoff:
groups[group_num].append(i)
else:
groups[num] = [num, i]
group_num = num
num += 1
else:
groups[num] = [num, i]
group_num = num
num += 1
last_node = node
for i in groups:
num = groups[i][0]
h = num * 360 / len(groups)
the_col = hsl_to_str(h, 0.5, 0.5)
style = NodeStyle()
style['size'] = 0
style["vt_line_color"] = the_col
style["hz_line_color"] = the_col
style["vt_line_width"] = 2
style["hz_line_width"] = 2
if len(groups[i]) == 2:
ca = t.search_nodes(name=groups[i][1])[0]
ca.set_style(style)
else:
ca = t.get_common_ancestor(groups[i][1:])
ca.set_style(style)
tocolor = []
for j in ca.children:
tocolor.append(j)
while len(tocolor) > 0:
x = tocolor.pop(0)
x.set_style(style)
for j in x.children:
tocolor.append(j)
ca_list.append((ca, h))
# for each common ancestor node get it's closest common ancestor neighbour and find the common ancestor of those two nodes
# colour the common ancestor then add it to the group - continue until only the root node is left
while len(ca_list) > 1:
distance = float('inf')
got_one = False
for i, col1 in ca_list:
for j, col2 in ca_list:
if not i is j:
parent = t.get_common_ancestor(i, j)
getit = True
for children in parent.children:
if children != i and children != j:
getit = False
break
if getit:
the_dist = t.get_distance(i, j)
if the_dist <= distance:
distance = the_dist
the_i = i
the_j = j
the_i_col = col1
the_j_col = col2
got_one = True
if not got_one:
break
ca_list.remove((the_i, the_i_col))
ca_list.remove((the_j, the_j_col))
new_col = (the_i_col + the_j_col) / 2
new_node = t.get_common_ancestor(the_i, the_j)
the_col = hsl_to_str(new_col, 0.5, 0.3)
style = NodeStyle()
style['size'] = 0
style["vt_line_color"] = the_col
style["hz_line_color"] = the_col
style["vt_line_width"] = 2
style["hz_line_width"] = 2
new_node.set_style(style)
ca_list.append((new_node, new_col))
# if you just want a black tree
else:
style = NodeStyle()
style['size'] = 0
style["vt_line_color"] = '#000000'
style["hz_line_color"] = '#000000'
style["vt_line_width"] = 1
style["hz_line_width"] = 1
for n in t.traverse():
n.set_style(style)
color_list = [(240, 163, 255), (0, 117, 220), (153, 63, 0), (76, 0, 92),
(25, 25, 25), (0, 92, 49), (43, 206, 72), (255, 204, 153),
(128, 128, 128), (148, 255, 181), (143, 124, 0),
(157, 204, 0), (194, 0, 136), (0, 51, 128), (255, 164, 5),
(255, 168, 187), (66, 102, 0), (255, 0, 16), (94, 241, 242),
(0, 153, 143), (224, 255, 102), (116, 10, 255), (153, 0, 0),
(255, 255, 128), (255, 255, 0), (255, 80, 5), (0, 0, 0),
(50, 50, 50)]
up_to_colour = {}
ts = TreeStyle()
column_list = []
width_dict = {}
if not grid_options is None:
colour_dict = {}
type_dict = {}
min_val_dict = {}
max_val_dict = {}
leaf_name_dict = {}
header_count = 0
the_columns = {}
if grid_options == 'auto':
with open(the_table) as f:
headers = f.readline().rstrip().split('\t')[1:]
for i in headers:
the_columns[i] = [i]
type_dict[i] = 'colour'
colour_dict[i] = {'empty': '#FFFFFF'}
width_dict[i] = 20
up_to_colour[i] = 0
column_list.append(i)
else:
with open(grid_options) as g:
for line in g:
if line.startswith('H'):
name, type, width = line.rstrip().split('\t')[1:]
if name in the_columns:
the_columns[name].append(name + '_' +
str(header_count))
else:
the_columns[name] = [
name + '_' + str(header_count)
]
width = int(width)
name = name + '_' + str(header_count)
header_count += 1
colour_dict[name] = {'empty': '#FFFFFF'}
type_dict[name] = type
width_dict[name] = width
column_list.append(name)
up_to_colour[name] = 0
min_val_dict[name] = float('inf')
max_val_dict[name] = 0
elif line.startswith('C'):
c_name, c_col = line.rstrip().split('\t')[1:]
if not c_col.startswith('#'):
c_col = colorstr(map(int, c_col.split(',')))
colour_dict[name][c_name] = c_col
val_dict = {}
with open(the_table) as f:
headers = f.readline().rstrip().split('\t')[1:]
column_no = {}
for num, i in enumerate(headers):
if i in the_columns:
column_no[num] = i
for line in f:
name = line.split('\t')[0]
leaf_name = None
for n in t.traverse():
if n.is_leaf():
if name.split('.')[0] in n.name:
leaf_name = n.name
if leaf_name is None:
continue
else:
leaf_name_dict[leaf_name] = name
vals = line.rstrip().split('\t')[1:]
if name in val_dict:
sys.exit('Duplicate entry found in table.')
else:
val_dict[name] = {}
for num, val in enumerate(vals):
if num in column_no and val != '':
for q in the_columns[column_no[num]]:
column_name = q
if type_dict[column_name] == 'colour':
val_dict[name][column_name] = val
if not val in colour_dict[column_name]:
colour_dict[column_name][val] = colorstr(
color_list[up_to_colour[column_name] %
len(color_list)])
up_to_colour[column_name] += 1
elif type_dict[column_name] == 'text':
val_dict[name][column_name] = val
elif type_dict[column_name] == 'colour_scale_date':
year, month, day = val.split('-')
year, month, day = int(year), int(month), int(
day)
the_val = datetime.datetime(
year, month, day, 0, 0,
0) - datetime.datetime(
1970, 1, 1, 0, 0, 0)
val_dict[name][
column_name] = the_val.total_seconds()
if the_val.total_seconds(
) < min_val_dict[column_name]:
min_val_dict[
column_name] = the_val.total_seconds()
if the_val.total_seconds(
) > max_val_dict[column_name]:
max_val_dict[
column_name] = the_val.total_seconds()
elif type_dict[column_name] == 'colour_scale':
the_val = float(val)
val_dict[name][column_name] = the_val
if the_val < min_val_dict[column_name]:
min_val_dict[column_name] = the_val
if the_val > max_val_dict[column_name]:
max_val_dict[column_name] = the_val
else:
sys.exit('Unknown column type')
if not out_file is None:
new_desc = open(out_file + '.new_desc', 'w')
else:
new_desc = open('viridis.new_desc', 'w')
ts.legend_position = 3
leg_column = 0
for num, i in enumerate(column_list):
nameF = TextFace(font_gap * ' ' + i.rsplit('_', 1)[0] +
' ' * font_buffer,
fsize=font_size,
ftype=font_type,
tight_text=True)
nameF.rotation = -90
ts.aligned_header.add_face(nameF, column=num + 1)
new_desc.write('H\t' + i.rsplit('_', 1)[0] + '\t' + type_dict[i] +
'\t' + str(width_dict[i]) + '\n')
x = num * 200
if type_dict[i] == 'colour':
ts.legend.add_face(TextFace(
font_gap * ' ' + i.rsplit('_', 1)[0] + ' ' * font_buffer,
fsize=font_size,
ftype=font_type,
tight_text=True),
column=leg_column + 1)
ts.legend.add_face(RectFace(width_dict[i], 20, '#FFFFFF',
'#FFFFFF'),
column=leg_column)
for num2, j in enumerate(colour_dict[i]):
new_desc.write('C\t' + j + '\t' + colour_dict[i][j] + '\n')
ts.legend.add_face(TextFace(font_gap * ' ' + j +
' ' * font_buffer,
fsize=font_size,
ftype=font_type,
tight_text=True),
column=leg_column + 1)
ts.legend.add_face(RectFace(width_dict[i], 20,
colour_dict[i][j],
colour_dict[i][j]),
column=leg_column)
leg_column += 2
elif type_dict[i] == 'colour_scale':
ts.legend.add_face(TextFace(
font_gap * ' ' + i.rsplit('_', 1)[0] + ' ' * font_buffer,
fsize=font_size,
ftype=font_type,
tight_text=True),
column=leg_column + 1)
ts.legend.add_face(RectFace(width_dict[i], 20, '#FFFFFF',
'#FFFFFF'),
column=leg_column)
for num2 in range(11):
y = num2 * 20 + 30
val = (max_val_dict[i] - min_val_dict[i]) * num2 / 10.0
h = val / (max_val_dict[i] - min_val_dict[i]) * 270
s = 0.5
l = 0.5
colour = hsl_to_str(h, s, l)
ts.legend.add_face(TextFace(font_gap * ' ' + str(val) +
' ' * font_buffer,
fsize=font_size,
ftype=font_type,
tight_text=True),
column=leg_column + 1)
ts.legend.add_face(RectFace(width_dict[i], 20, colour,
colour),
column=leg_column)
leg_column += 2
elif type_dict[i] == 'colour_scale_date':
ts.legend.add_face(TextFace(
font_gap * ' ' + i.rsplit('_', 1)[0] + ' ' * font_buffer,
fsize=font_size,
ftype=font_type,
tight_text=True),
column=leg_column + 1)
ts.legend.add_face(RectFace(width_dict[i], 20, '#FFFFFF',
'#FFFFFF'),
column=leg_column)
for num2 in range(11):
y = num2 * 20 + 30
val = (max_val_dict[i] - min_val_dict[i]) * num2 / 10.0
h = val / (max_val_dict[i] - min_val_dict[i]) * 360
s = 0.5
l = 0.5
colour = hsl_to_str(h, s, l)
days = str(int(val / 60 / 60 / 24)) + ' days'
ts.legend.add_face(TextFace(font_gap * ' ' + days +
' ' * font_buffer,
fsize=font_size,
ftype=font_type,
tight_text=True),
column=leg_column + 1)
ts.legend.add_face(RectFace(width_dict[i], 20, colour,
colour),
column=leg_column)
leg_column += 2
for n in t.traverse():
if n.is_leaf():
name = leaf_name_dict[n.name]
if i in val_dict[name]:
val = val_dict[name][i]
else:
val = 'empty'
if type_dict[i] == 'colour':
n.add_face(RectFace(width_dict[i], 20,
colour_dict[i][val],
colour_dict[i][val]),
column=num + 1,
position="aligned")
elif type_dict[i] == 'colour_scale' or type_dict[
i] == 'colour_scale_date':
if val == 'empty':
colour = '#FFFFFF'
else:
h = (val - min_val_dict[i]) / (
max_val_dict[i] - min_val_dict[i]) * 360
s = 0.5
l = 0.5
colour = hsl_to_str(h, s, l)
n.add_face(RectFace(width_dict[i], 20, colour, colour),
column=num + 1,
position="aligned")
elif type_dict[i] == 'text':
n.add_face(TextFace(font_gap * ' ' + val +
' ' * font_buffer,
fsize=font_size,
ftype=font_type,
tight_text=True),
column=num + 1,
position="aligned")
if not pres_abs is None:
starting_col = len(column_list) + 1
subprocess.Popen('makeblastdb -out tempdb -dbtype prot -in ' +
pres_abs[0],
shell=True).wait()
folder = pres_abs[1]
len_dict = {}
gene_list = []
ts.legend.add_face(TextFace(font_gap * ' ' + 'Gene present/absent' +
' ' * font_buffer,
fsize=font_size,
ftype=font_type,
tight_text=True),
column=starting_col + 1)
ts.legend.add_face(RectFace(20, 20, '#FFFFFF', '#FFFFFF'),
column=starting_col)
ts.legend.add_face(TextFace(font_gap * ' ' + 'Gene present/absent' +
' ' * font_buffer,
fsize=font_size,
ftype=font_type,
tight_text=True),
column=starting_col + 1)
ts.legend.add_face(RectFace(20, 20, "#5ba965", "#5ba965"),
column=starting_col)
ts.legend.add_face(TextFace(font_gap * ' ' + 'Gene present/absent' +
' ' * font_buffer,
fsize=font_size,
ftype=font_type,
tight_text=True),
column=starting_col + 1)
ts.legend.add_face(RectFace(20, 20, "#cb5b4c", "#cb5b4c"),
column=starting_col)
with open(pres_abs[0]) as f:
for line in f:
if line.startswith('>'):
name = line.split()[0][1:]
gene_list.append(name)
len_dict[name] = 0
nameF = TextFace(font_gap * ' ' + name + ' ' * font_buffer,
fsize=font_size,
ftype=font_type,
tight_text=True)
nameF.rotation = -90
ts.aligned_header.add_face(nameF,
column=starting_col +
len(gene_list) - 1)
else:
len_dict[name] += len(line.rstrip())
min_length = 0.9
min_ident = 90
for n in t.iter_leaves():
the_name = n.name
if the_name[0] == '"' and the_name[-1] == '"':
the_name = the_name[1:-1]
if the_name.endswith('.ref'):
the_name = the_name[:-4]
if not os.path.exists(folder + '/' + the_name):
for q in os.listdir(folder):
if q.startswith(the_name):
the_name = q
if not os.path.exists(the_name + '.blast'):
subprocess.Popen(
'blastx -query ' + folder + '/' + the_name +
' -db tempdb -outfmt 6 -num_threads 24 -out ' + the_name +
'.blast',
shell=True).wait()
gotit = set()
with open(the_name + '.blast') as b:
for line in b:
query, subject, ident, length = line.split()[:4]
ident = float(ident)
length = int(length)
if ident >= min_ident and length >= min_length * len_dict[
subject]:
gotit.add(subject)
for num, i in enumerate(gene_list):
if i in gotit:
colour = "#5ba965"
else:
colour = "#cb5b4c"
n.add_face(RectFace(20, 20, colour, colour),
column=num + starting_col,
position="aligned")
# for num, i in enumerate(gene_list):
# x = (starting_col + num) * 200
# svg.writeString(i, x+50, 20, 12)
# y = 30
# svg.drawOutRect(x + 50, y, 12, 12, strtorgb('#5ba965'), strtorgb('#5ba965'), lt=0)
# svg.writeString('present', x + 70, y + 12, 12)
# y = 50
# svg.drawOutRect(x + 50, y, 12, 12, strtorgb('#cb5b4c'), strtorgb('#cb5b4c'), lt=0)
# svg.writeString('absent', x + 70, y + 12, 12)
# Set these to False if you don't want bootstrap/distance values
ts.show_branch_length = label
ts.show_branch_support = bootstrap
ts.show_leaf_name = False
for node in t.traverse():
if node.is_leaf():
node.add_face(AttrFace("name",
fsize=font_size,
ftype=font_type,
tight_text=True,
fgcolor='black'),
column=0,
position="aligned")
ts.margin_left = 20
ts.margin_right = 100
ts.margin_top = 20
ts.margin_bottom = 20
if extend:
ts.draw_guiding_lines = True
ts.scale = the_scale
if not circular is None:
ts.mode = "c"
ts.arc_start = 0
ts.arc_span = 360
if out_file is None:
t.show(tree_style=ts)
else:
t.render(out_file, w=210, units='mm', tree_style=ts)
def create_circle(self,
filename,
title,
name2color=None,
name2class_dic=None,
class2color_dic=None,
vector=None,
ignore_branch_length=True):
plt.clf()
axis_font = {'size': '3'}
plt.rc('xtick', labelsize=0.1)
plt.rc('ytick', labelsize=0.1)
plt.rc({'font.size': 0.1})
# legend creation
if name2class_dic and class2color_dic:
leg = []
for cls, color in class2color_dic.items():
leg.append(mpatches.Patch(color=color, label=cls))
t = Tree(self.nwk)
# iterate over tree leaves only
for l in t.iter_leaves():
ns = NodeStyle()
if name2color:
ns["bgcolor"] = name2color[
l.name] if l.name in name2color else 'white'
elif name2class_dic and class2color_dic:
ns["bgcolor"] = class2color_dic[name2class_dic[
l.name]] if l.name in name2class_dic else 'white'
# Gray dashed branch lines
#ns["hz_line_type"] = 1
#ns["hz_line_color"] = "#cccccc"
#
l.img_style = ns
F = TextFace(l.name)
F.ftype = 'Times'
#if vector:
# if l.name in vector:
l.add_features(profile=[random.random()
for x in range(10)]) #vector[l.name])
l.add_features(deviation=[0 for x in range(10)
]) #len(vector[l.name]))])
l.add_face(ProfileFace(max_v=1,
min_v=0.0,
center_v=0.5,
width=200,
height=40,
style='heatmap',
colorscheme=5),
column=0,
position='aligned')
# Create an empty TreeStyle
ts = TreeStyle()
# Set our custom layout function
ts.layout_fn = VisualizeCircularTree.layout
# Draw a tree
ts.mode = "c"
# We will add node names manually
ts.show_leaf_name = False
# Show branch data
ts.show_branch_length = True
ts.show_branch_support = True
ts.force_topology = ignore_branch_length
ts.title.add_face(TextFace(title, fsize=20, ftype='Times'), column=15)
# legend creation
if name2class_dic and class2color_dic:
for k, (cls, col) in enumerate(class2color_dic.items()):
x = RectFace(8, 8, 'black', col)
#x.opacity=0.5
ts.legend.add_face(x, column=8)
ts.legend.add_face(TextFace(' ' + cls + ' ',
fsize=9,
ftype='Times'),
column=9)
t.render(filename + '.pdf', tree_style=ts, dpi=5000)
big_tree.ladderize()
ts = TreeStyle()
ts.show_leaf_name = False
#ts.show_branch_support = True
ts.scale = 100000
if mode == 'small':
ts.scale = 750000
#add legend
for each in list(colourDict.keys()):
ts.legend.add_face(CircleFace(radius=size[mode]/2, color=colourDict[each]), column=0)
ts.legend.add_face(TextFace(each, ftype="Helvetica", fsize=size[mode]), column=1)
ts.legend.add_face(CircleFace(radius=size[mode]/2, color='#F1F1F1'), column=0)
ts.legend.add_face(TextFace('Guinea', ftype="Helvetica", fsize=size[mode]), column=1)
ts.legend.add_face(RectFace(width=size[mode], height=size[mode], bgcolor='#D3D3D3', fgcolor='#FFFFFF'), column=0)
ts.legend.add_face(TextFace('Sierra Leone', ftype="Helvetica", fsize=size[mode]), column=1)
ts.legend.add_face(RectFace(triangle=True, width=size[mode], height=size[mode], bgcolor='#939393', fgcolor='#FFFFFF'), column=0)
ts.legend.add_face(TextFace('Liberia', ftype="Helvetica", fsize=size[mode]), column=1)
#reset nodes
ns = NodeStyle()
ns['size'] = 0
ns['hz_line_width'] = 2
ns['vt_line_width'] = 2
if mode =='big':
ns['hz_line_width'] = 0
ns['vt_line_width'] = 0
for n in big_tree.traverse():
n.set_style(ns)
def format( tree , genedict , detection , includenames , filename , speciescolors = None):
print 'final output...'
red = Color('red')
blue = Color('blue')
colorvec = list(red.range_to(blue, len(genedict.keys())))
colormap = {}
columnmap = {}
for i,ref in enumerate(genedict):
if ref not in detection:
columnmap[ref] = 3
if 'hybrid' in ref.lower():
columnmap[ref] = 0
if 'eff' in ref.lower():
columnmap[ref] = 1
if 'hap' in ref.lower():
columnmap[ref] = 2
colormap[ref] = colorvec[i].hex
for i,ref in enumerate(detection):
columnmap[ref] = 3 + i
colormap[ref] = colorvec[i].hex
print columnmap
print colormap
circledict = {}
for n in t.traverse():
nst = NodeStyle()
nst["size"] = 0
nst["fgcolor"] = 'black'
nst["hz_line_width"] = 4
nst["vt_line_width"]= 4
nst.show_name = False
if n.is_leaf():
if speciescolors != None and n.name in speciescolors:
nst["bgcolor"] = colors[n.name]
nst.show_name = True
n.add_face( AttrFace(attr = 'name', ftype='Helvetica', fgcolor='black', fsize =18 ,fstyle = 'normal' ), column =0 )
refs = json.loads(n.refs)
for ref in genedict:
if ref in refs and ref in detection:
n.add_face( CircleFace ( 10 , colormap[ref]), column = 2 + columnmap[ref] )
n.img_style = nst
if ref in refs and ref not in detection:
n.add_face( RectFace ( 20 , 20 , colormap[ref], colormap[ref] ), column = 2 + columnmap[ref] )
n.img_style = nst
if ref not in refs and ref not in detection:
n.add_face( RectFace ( 20 , 20 , colormap[ref], 'white' ), column = 2 + columnmap[ref] )
n.img_style = nst
###color by species
if n.name in speciescolors:
nst['bgcolor'] = speciescolors[n.name]
else:
if n.name.strip() in includenames:
n.add_face( AttrFace(attr = 'name', ftype='Helvetica', fgcolor='black', fsize =20 ,fstyle = 'normal' ), column =0 )
nst.size = 2
n.img_style = nst
else:
nst.size = 0
n.img_style = nst
ts = TreeStyle()
for i,ref in enumerate(colormap.keys()):
if 'ubi' not in ref:
ts.title.add_face(TextFace(ref, fsize=12), column=0)
ts.title.add_face( RectFace(10 , 10 , colormap[ref] , colormap[ref]), column = 1)
ts.show_leaf_name=False
"""ts.mode = "c"
ts.arc_start = 270
ts.arc_span = 359
ts.root_opening_factor = 1
"""
ts.scale = 190
t.show(tree_style = ts)
t.render(filename + ".png", tree_style = ts)
t.render( filename +".svg" , tree_style = ts)
if "coli" in node.sci_name:
circle = RectFace(100, 100, "red", "red")
faces.add_face_to_node(circle,
position="aligned",
column=1,
node=node)
if "Salm" in node.sci_name:
circle = RectFace(100, 100, "blue", "blue")
faces.add_face_to_node(circle,
position="aligned",
column=2,
node=node)
ts = TreeStyle()
ts.branch_vertical_margin = 10
ts.allow_face_overlap = False
ts.show_scale = False
ts.show_leaf_name = False
ts.layout_fn = layout
ts.mode = "c"
ts.root_opening_factor = .25
ts.optimal_scale_level = "full"
ts.force_topology = True
for i, j in enumerate(lineage_colors.keys()):
ts.legend.add_face(RectFace(10000, 300, lineage_colors[j],
lineage_colors[j]),
column=i)
ts.legend.add_face(TextFace(j, fsize=600), column=i)
tree.show(tree_style=ts)
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 traitTestTrees(scenarios, traits, mapper, outDir, treePath, floatSwitch=0):
# styles for convergent and transition nodes
styleConv = NodeStyle()
styleConv["fgcolor"] = "black"
styleConv["size"] = 0
styleConv["hz_line_width"] = branchWidth
styleConv["vt_line_width"] = branchWidth
styleConv["hz_line_color"] = "orange"
styleConv["vt_line_color"] = "orange"
styleTran = NodeStyle()
styleTran["fgcolor"] = "black"
styleTran["size"] = 0
styleTran["hz_line_width"] = branchWidth
styleTran["vt_line_width"] = branchWidth
styleTran["vt_line_color"] = "orange"
# face object indicating a transition
block = RectFace(4, 20, "blue", "blue")
block.margin_right = 2
def rgb2hex(r, g, b):
hex = "#{:02x}{:02x}{:02x}".format(r, g, b)
return hex
# loop thru trees
for cutoff in sorted(scenarios.keys()):
tree = init_tree(treePath)
#tree.convert_to_ultrametric(tree_length=1)
# parse scenario
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:])
# loop thru nodes
for n in tree.traverse():
#outline for the circle node
#ol = CircleFace(11, color = "black")#
#n.add_face(ol, column=1, position="float-behind") #float-behind
# draw circle face with mapped color
nd = CircleFace(
14, color=rgb2hex(*[int(val) for val in mapper(traits[n.ND])]))
nd.background.color = None
nd.border.width = None
if manual_mode_nodes:
if n.ND in manual_mode_nodes["T"]:
n.set_style(styleTran)
n.add_face(block, column=0, position="float")
elif n.ND in manual_mode_nodes["C"]:
n.set_style(styleConv)
else:
n.set_style(nstyle)
else:
nd.background.color = "white"
n.add_face(nd, column=2, position="float") # float
# this is necessary for some reason to keep the tree from collapsing
n.add_face(TextFace(" "), column=2, position="branch-bottom")
n.img_style["size"] = 0
outFile = str(round(cutoff, floatSwitch)).replace('.', '_') + ".pdf"
# prepend path to filename
outFile = outDir + '/' + outFile
tree.render(outFile, h=6, units="in", tree_style=tree_style)
print >> sys.stderr, outFile
def draw_tree(variant_dict, profile_dict, the_tree, ref_name):
major_alleles, minor_alleles = profile_dict
t = Tree(the_tree, quoted_node_names=True)
font_size = 8
font_type = 'Heveltica'
font_gap = 3
font_buffer = 10
ts = TreeStyle()
position_list = None
allele_count = {}
max_major = 0
max_minor = 0
max_combo = 0
max_major_minor = 0
for n in t.iter_leaves():
the_name = n.name
if the_name == ref_name:
t.set_outgroup(n)
if position_list is None:
position_list = list(variant_dict[the_name])
position_list.sort()
for num, i in enumerate(position_list):
nameF = TextFace(font_gap * ' ' + str(i) + ' ' * font_buffer,
fsize=font_size,
ftype=font_type,
tight_text=True)
nameF.rotation = -90
ts.aligned_header.add_face(nameF, column=num)
minor_allele, major_allele, missing = 0, 0, 0
for num, i in enumerate(position_list):
if i in major_alleles and variant_dict[the_name][
i] in major_alleles[i]:
s, l = 0.0, 0.1
major_allele += 1
elif i in minor_alleles and variant_dict[the_name][
i] in minor_alleles[i]:
s, l = 0.0, 0.5
minor_allele += 1
else:
s, l = 0.0, 0.9
missing += 1
if variant_dict[the_name][i] == 'a':
colour = colorstr(hsl_to_rgb(0, s, l))
#colour = "#65dad0"
elif variant_dict[the_name][i] == 't':
colour = colorstr(hsl_to_rgb(140, s, l))
#colour = "#daa3dc"
elif variant_dict[the_name][i] == 'c':
colour = colorstr(hsl_to_rgb(220, s, l))
#colour = "#9bd686"
elif variant_dict[the_name][i] == 'g':
colour = colorstr(hsl_to_rgb(300, s, l))
#colour = "#e1b86f"
else:
colour = "#ffffff"
#colour = colorstr(hsl_to_rgb(h, s, l))
n.add_face(RectFace(20, 20, colour, colour),
column=num,
position="aligned")
allele_count[the_name] = (major_allele, minor_allele, missing)
if major_allele > max_major:
max_major = major_allele
if minor_allele + major_allele >= max_combo:
max_combo = minor_allele + major_allele
if minor_allele >= max_minor:
max_minor = minor_allele
if major_allele > max_major_minor:
max_major_minor = major_allele
nst1 = NodeStyle()
nst1["bgcolor"] = "LightSteelBlue"
nst2 = NodeStyle()
nst2["bgcolor"] = "Moccasin"
with open(args.out_text, 'w') as o:
for n in t.iter_leaves():
the_name = n.name
major_allele, minor_allele, missing = allele_count[the_name]
if major_allele == max_major:
o.write("max_major\t%s\t%d\t%d\t%d\n" %
(the_name, major_allele, minor_allele, missing))
n.img_style["bgcolor"] = "#cb6a49"
elif minor_allele == max_minor and major_allele == max_major_minor:
o.write("max_minor\t%s\t%d\t%d\t%d\n" %
(the_name, major_allele, minor_allele, missing))
n.img_style["bgcolor"] = "#7aa457"
elif minor_allele + major_allele == max_combo:
o.write("max_combo\t%s\t%d\t%d\t%d\n" %
(the_name, major_allele, minor_allele, missing))
n.img_style["bgcolor"] = "#a46cb7"
n.add_face(TextFace('%d/%d/%d' %
(major_allele, minor_allele, missing),
fsize=font_size,
ftype=font_type,
tight_text=True),
column=len(position_list),
position="aligned")
ts.legend.add_face(TextFace('A',
fsize=font_size,
ftype=font_type,
tight_text=True),
column=0)
s = 0.5
ts.legend.add_face(RectFace(20, 20, colorstr(hsl_to_rgb(0, s, 0.3)),
colorstr(hsl_to_rgb(0, s, 0.3))),
column=1)
ts.legend.add_face(RectFace(20, 20, colorstr(hsl_to_rgb(0, s, 0.5)),
colorstr(hsl_to_rgb(0, s, 0.5))),
column=2)
ts.legend.add_face(RectFace(20, 20, colorstr(hsl_to_rgb(0, s, 0.8)),
colorstr(hsl_to_rgb(0, s, 0.8))),
column=3)
ts.legend.add_face(TextFace('T',
fsize=font_size,
ftype=font_type,
tight_text=True),
column=0)
ts.legend.add_face(RectFace(20, 20, colorstr(hsl_to_rgb(140, s, 0.3)),
colorstr(hsl_to_rgb(140, s, 0.3))),
column=1)
ts.legend.add_face(RectFace(20, 20, colorstr(hsl_to_rgb(140, s, 0.5)),
colorstr(hsl_to_rgb(140, s, 0.5))),
column=2)
ts.legend.add_face(RectFace(20, 20, colorstr(hsl_to_rgb(140, s, 0.8)),
colorstr(hsl_to_rgb(140, s, 0.8))),
column=3)
ts.legend.add_face(TextFace('C',
fsize=font_size,
ftype=font_type,
tight_text=True),
column=0)
ts.legend.add_face(RectFace(20, 20, colorstr(hsl_to_rgb(220, s, 0.3)),
colorstr(hsl_to_rgb(220, s, 0.3))),
column=1)
ts.legend.add_face(RectFace(20, 20, colorstr(hsl_to_rgb(220, s, 0.5)),
colorstr(hsl_to_rgb(220, s, 0.5))),
column=2)
ts.legend.add_face(RectFace(20, 20, colorstr(hsl_to_rgb(220, s, 0.8)),
colorstr(hsl_to_rgb(220, s, 0.8))),
column=3)
ts.legend.add_face(TextFace('G',
fsize=font_size,
ftype=font_type,
tight_text=True),
column=0)
ts.legend.add_face(RectFace(20, 20, colorstr(hsl_to_rgb(300, s, 0.3)),
colorstr(hsl_to_rgb(300, s, 0.3))),
column=1)
ts.legend.add_face(RectFace(20, 20, colorstr(hsl_to_rgb(300, s, 0.5)),
colorstr(hsl_to_rgb(300, s, 0.5))),
column=2)
ts.legend.add_face(RectFace(20, 20, colorstr(hsl_to_rgb(300, s, 0.8)),
colorstr(hsl_to_rgb(300, s, 0.8))),
column=3)
ts.legend.add_face(TextFace('-',
fsize=font_size,
ftype=font_type,
tight_text=True),
column=0)
ts.legend.add_face(RectFace(20, 20, "#cccccc", "#cccccc"), column=1)
t.render(args.out_file, w=210, units='mm', tree_style=ts)
for output_string, mode, layout_fn in zip(
[
"tree_symbols",
"tree_symbols",
"tree_arrows",
], ["c", "r", "r"],
[botstrap_lower_right, botstrap_lower_right, botstrap_symbols]):
stars_style = TreeStyle()
stars_style.layout_fn = layout_fn
stars_style.orientation = tree_orientation
stars_style.mode = mode
for subtype in subtype_color_dict:
stars_style.legend.add_face(RectFace(10, 10,
subtype_color_dict[subtype],
subtype_color_dict[subtype]),
column=0)
stars_style.legend.add_face(TextFace(
subtype, fgcolor=subtype_color_dict[subtype], bold=True),
column=1)
stars_style.legend_position = 2
ref_tree.render(file_name="plots/{}_{}_{}.pdf".format(
plot_prefix, output_string, mode),
tree_style=stars_style)
print("Done.")
# dataframe for bootstrap values plotting
shared_edge_df = pd.DataFrame(shared_edge_support_values)
# "top-right" of scatter plot
high_support_df = shared_edge_df[
def draw_tree(tree, conf, outfile):
try:
from ete3 import (add_face_to_node, AttrFace, TextFace, TreeStyle, RectFace, CircleFace,
SequenceFace, random_color, SeqMotifFace)
except ImportError as e:
print(e)
return
def ly_basic(node):
if node.is_leaf():
node.img_style['size'] = 0
else:
node.img_style['size'] = 0
node.img_style['shape'] = 'square'
if len(MIXED_RES) > 1 and hasattr(node, "tree_seqtype"):
if node.tree_seqtype == "nt":
node.img_style["bgcolor"] = "#CFE6CA"
ntF = TextFace("nt", fsize=6, fgcolor='#444', ftype='Helvetica')
add_face_to_node(ntF, node, 10, position="branch-bottom")
if len(NPR_TREES) > 1 and hasattr(node, "tree_type"):
node.img_style['size'] = 4
node.img_style['fgcolor'] = "steelblue"
node.img_style['hz_line_width'] = 1
node.img_style['vt_line_width'] = 1
def ly_leaf_names(node):
if node.is_leaf():
spF = TextFace(node.species, fsize=10, fgcolor='#444444', fstyle='italic', ftype='Helvetica')
add_face_to_node(spF, node, column=0, position='branch-right')
if hasattr(node, 'genename'):
geneF = TextFace(" (%s)" %node.genename, fsize=8, fgcolor='#777777', ftype='Helvetica')
add_face_to_node(geneF, node, column=1, position='branch-right')
def ly_supports(node):
if not node.is_leaf() and node.up:
supFace = TextFace("%0.2g" %(node.support), fsize=7, fgcolor='indianred')
add_face_to_node(supFace, node, column=0, position='branch-top')
def ly_tax_labels(node):
if node.is_leaf():
c = LABEL_START_COL
largest = 0
for tname in TRACKED_CLADES:
if hasattr(node, "named_lineage") and tname in node.named_lineage:
linF = TextFace(tname, fsize=10, fgcolor='white')
linF.margin_left = 3
linF.margin_right = 2
linF.background.color = lin2color[tname]
add_face_to_node(linF, node, c, position='aligned')
c += 1
for n in range(c, len(TRACKED_CLADES)):
add_face_to_node(TextFace('', fsize=10, fgcolor='slategrey'), node, c, position='aligned')
c+=1
def ly_full_alg(node):
pass
def ly_block_alg(node):
if node.is_leaf():
if 'sequence' in node.features:
seqFace = SeqMotifFace(node.sequence, [])
# [10, 100, "[]", None, 10, "black", "rgradient:blue", "arial|8|white|domain Name"],
motifs = []
last_lt = None
for c, lt in enumerate(node.sequence):
if lt != '-':
if last_lt is None:
last_lt = c
if c+1 == len(node.sequence):
start, end = last_lt, c
motifs.append([start, end, "()", 0, 12, "slategrey", "slategrey", None])
last_lt = None
elif lt == '-':
if last_lt is not None:
start, end = last_lt, c-1
motifs.append([start, end, "()", 0, 12, "grey", "slategrey", None])
last_lt = None
seqFace = SeqMotifFace(node.sequence, motifs,
intermotif_format="line",
seqtail_format="line", scale_factor=ALG_SCALE)
add_face_to_node(seqFace, node, ALG_START_COL, aligned=True)
TRACKED_CLADES = ["Eukaryota", "Viridiplantae", "Fungi",
"Alveolata", "Metazoa", "Stramenopiles", "Rhodophyta",
"Amoebozoa", "Crypthophyta", "Bacteria",
"Alphaproteobacteria", "Betaproteobacteria", "Cyanobacteria",
"Gammaproteobacteria",]
# ["Opisthokonta", "Apicomplexa"]
colors = random_color(num=len(TRACKED_CLADES), s=0.45)
lin2color = dict([(ln, colors[i]) for i, ln in enumerate(TRACKED_CLADES)])
NAME_FACE = AttrFace('name', fsize=10, fgcolor='#444444')
LABEL_START_COL = 10
ALG_START_COL = 40
ts = TreeStyle()
ts.draw_aligned_faces_as_table = False
ts.draw_guiding_lines = False
ts.show_leaf_name = False
ts.show_branch_support = False
ts.scale = 160
ts.layout_fn = [ly_basic, ly_leaf_names, ly_supports, ly_tax_labels]
MIXED_RES = set()
MAX_SEQ_LEN = 0
NPR_TREES = []
for n in tree.traverse():
if hasattr(n, "tree_seqtype"):
MIXED_RES.add(n.tree_seqtype)
if hasattr(n, "tree_type"):
NPR_TREES.append(n.tree_type)
seq = getattr(n, "sequence", "")
MAX_SEQ_LEN = max(len(seq), MAX_SEQ_LEN)
if MAX_SEQ_LEN:
ALG_SCALE = min(1, 1000./MAX_SEQ_LEN)
ts.layout_fn.append(ly_block_alg)
if len(NPR_TREES) > 1:
rF = RectFace(4, 4, "steelblue", "steelblue")
rF.margin_right = 10
rF.margin_left = 10
ts.legend.add_face(rF, 0)
ts.legend.add_face(TextFace(" NPR node"), 1)
ts.legend_position = 3
if len(MIXED_RES) > 1:
rF = RectFace(20, 20, "#CFE6CA", "#CFE6CA")
rF.margin_right = 10
rF.margin_left = 10
ts.legend.add_face(rF, 0)
ts.legend.add_face(TextFace(" Nucleotide based alignment"), 1)
ts.legend_position = 3
try:
tree.set_species_naming_function(spname)
annotate_tree_with_ncbi(tree)
a = tree.search_nodes(species='Dictyostelium discoideum')[0]
b = tree.search_nodes(species='Chondrus crispus')[0]
#out = tree.get_common_ancestor([a, b])
#out = tree.search_nodes(species='Haemophilus parahaemolyticus')[0].up
tree.set_outgroup(out)
tree.swap_children()
except Exception:
pass
tree.render(outfile, tree_style=ts, w=170, units='mm', dpi=150)
tree.render(outfile+'.svg', tree_style=ts, w=170, units='mm', dpi=150)
tree.render(outfile+'.pdf', tree_style=ts, w=170, units='mm', dpi=150)
def create_circle(self, filename, title, ignore_branch_length=True):
plt.clf()
axis_font = {'size': '29'}
plt.rc('xtick', labelsize=0.1)
plt.rc('ytick', labelsize=0.1)
plt.rc({'font.size': 0.2})
plt.rc('text', usetex=True)
# legend creation
if self.motif2struct and self.struct2color_dic:
leg = []
for cls, color in self.struct2color_dic.items():
leg.append(mpatches.Patch(color=color, label=str(cls)))
t = Tree(self.nwk)
node_to_keep = []
for l in t.iter_leaves():
if len(l.name) > 1:
node_to_keep.append(l.name)
t.prune(tuple(node_to_keep))
# iterate over tree leaves only
for l in t.iter_leaves():
ns = NodeStyle()
if self.motif2struct and self.struct2color_dic:
ns["bgcolor"] = self.struct2color_dic[self.motif2struct[
l.
name]] if l.name in self.motif2struct and self.motif2struct[
l.name] in self.struct2color_dic else 'white'
# Gray dashed branch lines
#ns["hz_line_type"] = 1
#ns["hz_line_color"] = "#cccccc"
#
l.img_style = ns
if self.propVec:
if l.name in self.propVec:
l.add_features(profile=self.propVec[l.name])
l.add_features(deviation=[
0 for x in range(len(self.propVec[l.name]))
])
l.add_face(ProfileFace(max_v=1,
min_v=-1,
center_v=0,
width=200,
height=40,
style='heatmap',
colorscheme=2),
column=3,
position='aligned')
l.name = " " + l.name + " "
# Create an empty TreeStyle
ts = TreeStyle()
# Set our custom layout function
ts.layout_fn = VisualizeTreeOfMotifs.layout
# Draw a tree
ts.mode = "c"
# We will add node names manually
ts.show_leaf_name = False
# Show branch data
ts.show_branch_length = False
ts.show_branch_support = False
ts.force_topology = ignore_branch_length
ts.title.add_face(TextFace(title, fsize=20, ftype='Times'), column=15)
# legend creation
if self.motif2struct and self.struct2color_dic:
for k, (cls, col) in enumerate(self.struct2color_dic.items()):
x = RectFace(15, 15, 'black', col)
#x.opacity=0.5
ts.legend.add_face(x, column=8)
ts.legend.add_face(TextFace(' ' + str(cls) + ' ',
fsize=30,
ftype='Times'),
column=9)
ts.legend.add_face(TextFace('--- Properties vector order ↓',
fsize=30,
ftype='Times'),
column=10)
for y in [
'Mean molecular weight of amino acids',
'Mean flexibility of amino acids',
'Mean DIWV instability index of sequence',
'Mean surface accessibility of amino acids',
'Mean KD hydrophobicity', 'Mean hydrophilicity of amino acids'
]:
x = RectFace(5, 5, 'black', 'black')
#x.opacity=0.5
ts.legend.add_face(x, column=11)
ts.legend.add_face(TextFace(' ' + y + ' ',
fsize=25,
ftype='Times'),
column=12)
t.render(filename + '.pdf', tree_style=ts, dpi=5000)
columnmap[fasta] = i
colormap[fasta] = colorvec[i].hex
annotated = []
print speciescolors
for fasta in genedict:
for leaf in t.get_leaves():
nst = NodeStyle()
nst["size"] = 0
nst["fgcolor"] = 'black'
nst["hz_line_width"] = 2
nst["vt_line_width"]= 2
nst.show_name = True
if leaf.name.split('/')[0] in genedict[fasta]:
if 'HH' not in fasta and 'LOMETS' not in fasta:
leaf.add_face( RectFace ( 10 , 10 , colormap[fasta], colormap[fasta] ), column = columnmap[fasta] )
if leaf not in annotated:
try:
face = leaf.add_face( TextFace ( text = genedict[fasta][leaf.name.split('/')[0]][2]) , column = 10 )
annotated.append(leaf)
except:
print genedict[fasta][leaf.name.split('/')[0]]
if colorSepcies == True:
lineage = genedict[fasta][leaf.name.split('/')[0]][3].split(';') + [genedict[fasta][leaf.name.split('/')[0]][2]] + genedict[fasta][leaf.name.split('/')[0]][2].split()
for taxa in lineage[-1:]:
if taxa in speciescolors:
print taxa
print speciescolors[taxa]
nst['bgcolor'] = speciescolors[taxa]
def draw_lifting_tree_inline(filename: str) -> None:
ts = TreeStyle()
ts.show_leaf_name = False
ts.layout_fn = layout_lift
ts.rotation = 90
ts.branch_vertical_margin = 10
ts.show_scale = False
ts.scale = 50
ts.title.add_face(TextFace(" ", fsize=20), column=0)
ts.legend.add_face(TextFace(" "), column=0)
ts.legend.add_face(TextFace(" "), column=1)
ts.legend.add_face(TextFace(" "), column=2)
ts.legend.add_face(TextFace(" "), column=3)
ts.legend.add_face(TextFace(" "), column=0)
ts.legend.add_face(TextFace("Node shape and size:"), column=1)
ts.legend.add_face(TextFace(" "), column=2)
ts.legend.add_face(TextFace("Node color - membership value:"), column=3)
ts.legend.add_face(TextFace(" "), column=0)
ts.legend.add_face(TextFace(" "), column=1)
ts.legend.add_face(TextFace(" "), column=2)
ts.legend.add_face(TextFace(" "), column=3)
ts.legend.add_face(PieChartFace([100], 20, 20, colors=['white'], line_color='black'), column=0)
ts.legend.add_face(TextFace(" topics that relate to the cluster"), column=1)
ts.legend.add_face(RectFace(30, 10, "#90ee90", "#90ee90"), column=2)
ts.legend.add_face(TextFace(" topic with minor membership 0<u(t)<=0.2"), column=3)
ts.legend.add_face(TextFace(" "), column=0)
ts.legend.add_face(TextFace(" "), column=1)
ts.legend.add_face(TextFace(" "), column=2)
ts.legend.add_face(TextFace(" "), column=3)
ts.legend.add_face(PieChartFace([100], 40, 40, colors=['white'], line_color='black'), column=0)
ts.legend.add_face(TextFace(" head subject"), column=1)
ts.legend.add_face(RectFace(30, 10, "green", "green"), column=2)
ts.legend.add_face(TextFace(u" topic with medium membership 0.2<u(t)<=0.4 "), column=3)
ts.legend.add_face(TextFace(" "), column=0)
ts.legend.add_face(TextFace(" "), column=1)
ts.legend.add_face(TextFace(" "), column=2)
ts.legend.add_face(TextFace(" "), column=3)
ts.legend.add_face(RectFace(20, 20, "black", "white"), column=0)
ts.legend.add_face(TextFace(" topics that don't refer to cluster "), \
column=1)
ts.legend.add_face(RectFace(30, 10, "#004000", "#004000"), column=2)
ts.legend.add_face(TextFace(" topic with high membership u(t)>0.4"), column=3)
ts.legend.add_face(TextFace(" "), column=0)
ts.legend.add_face(TextFace(" "), column=1)
ts.legend.add_face(TextFace(" "), column=2)
ts.legend.add_face(TextFace(" "), column=3)
ts.legend.add_face(TextFace(" "), column=1)
ts.legend.add_face(TextFace(" "), column=1)
ts.legend.add_face(RectFace(30, 10, "red", "red"), column=2)
ts.legend.add_face(TextFace(" topic with no membership (u(t)=0)"), column=3)
ts.legend_position = 3
ete3_desc = read_ete3_from_file(filename)
tree = Tree(ete3_desc, format=1)
# tree.show(tree_style=ts)
return tree.render("%%inline",tree_style=ts)
def heatmap_view(tree, orthologous_groups, save_dir):
"""Generates a heatmap of regulation states in all species."""
light_tree = copy.deepcopy(tree) # Tree copy for the light heatmap
# Heat map settings
rect_face_fgcolor = 'black'
locus_tag_len = max(len(gene.locus_tag) + 5
for ortho_grp in orthologous_groups
for gene in ortho_grp.genes)
rect_face_width = locus_tag_len * 8
light_rect_face_width = 20
rect_face_height = 20
rotation = 90
# Sort orthologous groups by the number of regulated genes in each group
orthologous_groups = filter_and_sort_orthologous_grps(orthologous_groups)
# For each species and its gene in each orthologous group, draw a rectangle
for node, light_node in zip(tree.get_leaves(), light_tree.get_leaves()):
for i, orthologous_grp in enumerate(orthologous_groups, start=1):
#get all orthologs in group
matching_genes = [g for g in orthologous_grp.genes \
if g.genome.strain_name == node.name]
#if there is ortholog
if len(matching_genes) > 0:
# Get the first ortholog from the genome in the group
#this is the one with higher probability of regulation.
#so this probability will be displayed for the group
gene = matching_genes[0]
p_regulation = gene.operon.regulation_probability
p_notregulation = 1.0 - p_regulation
p_absence = 0
# No ortholog from this genome
else:
gene = None
p_regulation = 0
p_notregulation = 0
p_absence = 1
# Color of the rectangle is based on probabilities
rect_face_bgcolor = rgb2hex(
p_notregulation, p_regulation, p_absence)
rect_face_text = ('%s [%d]' % (gene.locus_tag, gene.operon.operon_id)
if gene else '')
rect_face_label = {'text': rect_face_text,
'font': 'Courier',
'fontsize': 8,
'color': 'black'}
# Create the rectangle
rect_face = RectFace(rect_face_width, rect_face_height,
rect_face_fgcolor, rect_face_bgcolor,
label=rect_face_label)
light_rect_face = RectFace(light_rect_face_width, rect_face_height,
rect_face_fgcolor, rect_face_bgcolor,
label='')
rect_face.rotation = -rotation
light_rect_face.rotation = -rotation
# Add the rectangle to the corresponding column
node.add_face(rect_face, column=i, position='aligned')
light_node.add_face(light_rect_face, column=i, position='aligned')
ts = TreeStyle()
# Add orthologous group descriptions
descriptions = ['-'.join([grp.description, str(grp.NOGs)]) for grp in orthologous_groups]
max_description_len = max(map(len, descriptions))
descriptions = [
'[%d]' % i + description + ' '*(max_description_len-len(description))
for i, description in enumerate(descriptions, start=1)]
for i, description in enumerate(descriptions, start=1):
text_face = TextFace(description, ftype='Courier')
text_face.hz_align = 1
text_face.vt_align = 1
text_face.rotation = -rotation
ts.aligned_header.add_face(text_face, column=i)
# Rotate the generated heatmap.
ts.margin_left = 10
ts.margin_top = 20
ts.rotation = rotation
ts.show_scale = False
# For some reason, it can't render to PDF in color
tree.render(os.path.join(save_dir, 'heatmap.svg'), tree_style=ts)
light_tree.render(os.path.join(save_dir, 'heatmap_light.svg'), tree_style=ts)
def 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)