def __init__(self, tree=None, length_attr="length", *args, **kwargs):
"""
Parameters
----------
tree : either a PhyloNode or TreeNode instance
length_attr : str
name of the attribute to use for length, defaults to 'length'
"""
if tree is not None:
children = [
type(self)(child, *args, **kwargs) for child in tree.children
]
PhyloNode.__init__(self,
params=tree.params.copy(),
children=children,
name=tree.name)
else:
PhyloNode.__init__(self, **kwargs)
# todo do we need to validate the length_attr key exists?
self._length = length_attr
self._node_space = 1.3
self._start = None
self._end = None
self._y = None
self._x = None
self._tip_rank = None
self._max_x = 0
self._min_x = 0
self._max_y = 0
self._min_y = 0
self._theta = 0
self._num_tips = 1
def test_missing_tip_name(self):
"""DndParser should produce the correct tree when missing a name"""
obs = DndParser(missing_tip_name)
exp = PhyloNode()
exp.append(PhyloNode())
exp.append(PhyloNode())
exp.children[0].append(PhyloNode(name="a"))
exp.children[0].append(PhyloNode(name="b"))
exp.children[1].append(PhyloNode(name="c"))
exp.children[1].append(PhyloNode())
self.assertEqual(str(obs), str(exp))
def test_nonames(self):
"""DndParser should produce the correct tree when there are no names"""
obs = DndParser(no_names)
exp = PhyloNode()
exp.append(PhyloNode())
exp.append(PhyloNode())
exp.children[0].append(PhyloNode())
exp.children[0].append(PhyloNode())
exp.children[1].append(PhyloNode())
exp.children[1].append(PhyloNode())
self.assertEqual(str(obs), str(exp))
def condense_node_order(matrix, smallest_index, node_order):
"""condenses two nodes in node_order based on smallest_index info
This function is used to create a tree while condensing a matrix
with the condense_matrix function. The smallest_index is retrieved
with find_smallest_index. The first index is replaced with a node object
that combines the two nodes corresponding to the indices in node order.
The second index in smallest_index is replaced with None.
Also sets the branch length of the nodes to 1/2 of the distance between
the nodes in the matrix"""
index1, index2 = smallest_index
node1 = node_order[index1]
node2 = node_order[index2]
# get the distance between the nodes and assign 1/2 the distance to the
# lengthproperty of each node
distance = matrix[index1, index2]
nodes = [node1, node2]
d = distance / 2.0
for n in nodes:
if n.children:
n.length = d - n.children[0].TipLength
else:
n.length = d
n.TipLength = d
# combine the two nodes into a new PhyloNode object
new_node = PhyloNode()
new_node.children.append(node1)
new_node.children.append(node2)
node1.parent = new_node
node2.parent = new_node
# replace the object at index1 with the combined node
node_order[index1] = new_node
# replace the object at index2 with None
node_order[index2] = None
return node_order
def test_gops(self):
"""Basic PhyloNode operations should work as expected"""
p = PhyloNode()
self.assertEqual(str(p), ";")
p.name = "abc"
self.assertEqual(str(p), "abc;")
p.length = 3
self.assertEqual(str(p), "abc:3;") # don't suppress branch from root
q = PhyloNode()
p.append(q)
self.assertEqual(str(p), "()abc:3;")
r = PhyloNode()
q.append(r)
self.assertEqual(str(p), "(())abc:3;")
r.name = "xyz"
self.assertEqual(str(p), "((xyz))abc:3;")
q.length = 2
self.assertEqual(str(p), "((xyz):2)abc:3;")
def test_minimal(self):
"""DndParser should produce the correct minimal tree"""
obs = DndParser(minimal)
exp = PhyloNode()
exp.append(PhyloNode())
self.assertEqual(str(obs), str(exp))
def get_tree(self, just_members=True):
"""returns the gene tree with tip names as gene stableid's
Arguments:
----------
- just_members: limits tips to just members of self
"""
gtrn = self.compara.ComparaDb.get_table("gene_tree_node")
condition = gtrn.c.root_id == list(self._gene_tree_root)[0]
query = sql.select(
[
gtrn.c.node_id,
gtrn.c.parent_id,
gtrn.c.distance_to_parent,
gtrn.c.seq_member_id,
],
whereclause=condition,
)
records = query.execute().fetchall()
# get the gene stable IDs, via join of seq_member with gene_member
# on seq_member_id
seqmem_ids = [r["seq_member_id"] for r in records]
seqmem = self.compara.ComparaDb.get_table("seq_member")
genmem = self.compara.ComparaDb.get_table("gene_member")
condition = genmem.c.gene_member_id == seqmem.c.gene_member_id
joined = seqmem.join(
genmem, genmem.c.gene_member_id == seqmem.c.gene_member_id)
query = (sql.select([
joined.c.seq_member_seq_member_id, joined.c.gene_member_stable_id
]).where(joined.c.seq_member_seq_member_id.in_(
seqmem_ids)).select_from(joined))
gene_ids = dict(query.execute().fetchall())
nodes = {}
parents = defaultdict(list)
for record in records:
parent_id = record["parent_id"]
node_id = record["node_id"]
length = record["distance_to_parent"]
name = gene_ids.get(record["seq_member_id"], None)
node = PhyloNode(length=length, name=name)
nodes[node_id] = node
parents[parent_id].append(node)
root = None
for parent in parents:
if parent not in nodes:
node = PhyloNode(name="root")
nodes[parent] = node
node = nodes[parent]
for child in parents[parent]:
child.parent = node
if len(parents[parent]) == 1:
root = node
if just_members:
stableids = [g.stableid for g in self.members]
root = root.get_sub_tree(stableids)
return root
def get_species_tree(self, just_members=True):
"""returns the species tree
Arguments:
----------
- just_members: limits tips to just members of self
"""
# grab the Ensembl species tree root ID
sptr = self.ComparaDb.get_table("species_tree_root")
condition = sptr.c.label == "Ensembl"
query = sql.select([sptr.c.root_id], whereclause=condition)
records = query.execute().fetchall()
assert len(records) == 1, records
root_id = records[0]["root_id"]
# get the tree nodes
sptn = self.ComparaDb.get_table("species_tree_node")
condition = sql.and_(sptn.c.root_id == root_id)
query = sql.select([sptn], whereclause=condition)
records = query.execute().fetchall()
# get the genome db -> name map
gen_db = self.ComparaDb.get_table("genome_db")
db_ids = [r["genome_db_id"] for r in records]
query = sql.select(
[gen_db.c.genome_db_id, gen_db.c.name],
whereclause=gen_db.c.genome_db_id.in_(db_ids),
)
id_name = dict(query.execute().fetchall())
for id_, name in id_name.items():
name = _Species.get_species_name(name)
id_name[id_] = None if name == "None" else name
nodes = {}
parents = defaultdict(list)
for record in records:
parent_id = record["parent_id"]
node_id = record["node_id"]
length = record["distance_to_parent"]
name = record["node_name"]
gen_dbid = record["genome_db_id"]
n = None if gen_dbid is None else id_name[gen_dbid]
name = name if n is None else n
node = PhyloNode(length=length, name=name)
nodes[node_id] = node
parents[parent_id].append(node)
root = None
for parent, value in parents.items():
if parent not in nodes:
node = PhyloNode(name="root")
nodes[parent] = node
node = nodes[parent]
for child in parents[parent]:
child.parent = node
if len(value) == 1:
root = node
# convert tip-names to match genome db names
if just_members:
root = root.get_sub_tree(self.species, tipsonly=True)
return root
