def write(obj, handle, **kwargs):
"""Write a new Nexus file containing the given trees.
Uses a simple Nexus template and the NewickIO writer to serialize just the
trees and minimal supporting info needed for a valid Nexus file.
"""
trees = list(obj)
writer = NewickIO.Writer(trees)
nexus_trees = [
TREE_TEMPLATE % {
'index': idx + 1,
'tree': nwk
} for idx, nwk in enumerate(
writer.to_strings(plain=False, plain_newick=True, **kwargs))
]
tax_labels = [
str(x.name) for x in chain(*(t.get_terminals() for t in trees))
]
text = NEX_TEMPLATE % {
'count': len(tax_labels),
'labels': ' '.join(tax_labels),
'trees': '\n'.join(nexus_trees),
}
handle.write(text)
return len(nexus_trees)
def nexus_text(obj, colour_branches, colours, **kwargs):
""" Take tree-like object(s) and create nexus-format representation.
Allows for colouring tip names.
Modified from http://biopython.org/DIST/docs/api/Bio.Phylo.NexusIO-pysrc.html
NB here we compensate for an apparent bug in the Biopython implementation,
whereby an additional colon is wrongly added to confidence values in the output tree strings.
"""
try:
trees = list(obj) # assume iterable
except TypeError:
trees = [obj]
writer = NewickIO.Writer(trees)
nexus_trees = [TREE_TEMPLATE % {'index': idx + 1, 'tree': nwk}
for idx, nwk in enumerate(
writer.to_strings(plain=False, plain_newick=True,
**kwargs))]
# if branches are being coloured, then taxon names already contain colouring annotation
# otherwise we need to add this annotation here
tax_labels = [ colour_taxon(str(x.name), colours) if not colour_branches else str(x.name) for x in chain(*(t.get_terminals() for t in trees))]
text = NEX_TEMPLATE % {
'count': len(tax_labels),
'labels': ' '.join(tax_labels), # taxlabels all on one line
'trees': '\n'.join(nexus_trees), # trees on separate lines
}
return re.sub(r':([0-9]{1,3}\.[0-9]{1,3}):', r'\1:', text) # Corrects for biopython bug. eg ":50.00:" -> "50.00:"
def to_nexus(self, filename):
"""Writes the tree to the given file in nexus format.
This method doesn't call Bio.Phylo.NexusIO as BayesTraitsV2 requires a
different dialect of Nexus format.
"""
# Copy the tree before making any changes on it.
tree = copy.deepcopy(self.tree)
# BayesTraits requires the Nexus file to have a "Translate" block which
# declares a number->taxon mapping so that numbers, not long taxa
# names, are used in the tree descriptions.
names_to_ints = dict((clade.name, i) for i, clade in enumerate(
tree.get_terminals(), start=1))
# Assign numbers to terminal clades
for node in tree.get_terminals():
node.name = str(names_to_ints[node.name])
# Drop names of the inner nodes
for n in tree.get_nonterminals():
n.name = None
# Tree to string
writer = NewickIO.Writer([tree])
nexus_tree = NEX_TEMPLATE % {
'translate': ',\n'.join('%d %s' % (name, id)
for id, name in names_to_ints.items()),
'tree': next(writer.to_strings(plain=False, plain_newick=True))}
# Write string to file
with open(filename, 'w') as handle:
handle.write(nexus_tree)
def write_nexus_trees_to_bayestraits(nx, handle, **kwargs):
"""Modified from Bio.Phylo.NexusIO.write():
add a translate block converting leaf names to integers.
"""
trees = BioNexusTrees_to_BioPhylo(nx.trees)
writer = NewickIO.Writer(trees)
nexus_trees = [
TREE_TEMPLATE % {"index": idx + 1, "tree": nwk}
for idx, nwk in enumerate(
writer.to_strings(plain=False, plain_newick=True, **kwargs)
)
]
translate = ["%d %s" % id_name for id_name in nx.translate.items()]
# Unused in my output format (BayesTraits) + why aren't they unique?
tax_labels = [taxon for nt in nx.trees for taxon in nt.get_taxa()]
#tax_labels = [str(x.name) for x in chain(*(t.get_terminals() for t in trees))]
text = NEX_TEMPLATE % {
"count": len(tax_labels),
"labels": " ".join(tax_labels),
"trees": "\n".join(nexus_trees),
"translate": ",\n ".join(translate)
}
handle.write(text)
return len(nexus_trees)
import sys
from Bio import Phylo
from Bio.Phylo import NewickIO
trees = list(Phylo.parse(sys.argv[1], "newick"))
print("Removing trees that are not bifurcating.")
for tree in trees:
for nonterminal in tree.get_nonterminals():
nonterminal.comment = None
nonterminal.branch_length = None
writer = NewickIO.Writer([tree for tree in trees if tree.is_bifurcating()])
print()
print("Saving trees as plain newick files (no branch lengths).")
with open(sys.argv[2], "w") as handle:
for newick_tree in writer.to_strings(plain=True):
handle.write(newick_tree + "\n")
