How to collapse leaves with the same name in ETE3 tree?

Question

[Andrzej Zielezinski]

I'm trying to collapse leaves that have the same name in ETE3's tree. I think I almost got it. Two leaves in my example (A. thaliana) have been combined into a single leaf, but not named. I would like it to have the same name as its components.

from ete3 import Tree

def collapsed_leaf(node):
    if len(node2labels[node]) == 1:
       return True
    else:
       return False


t1 = Tree('(("A. thaliana":0.723274,"A. thaliana":0.567784):0.067192,("B":0.279326,"H":0.756049):0.807788);', quoted_node_names=True)
node2labels = t1.get_cached_content(store_attr="name")
print(t1)
#       /-A. thaliana
#    /-|
#   |   \-A. thaliana
# --|
#   |   /-B
#    \-|
#       \-H

t2 = Tree(t1.write(is_leaf_fn=collapsed_leaf, quoted_node_names=True), quoted_node_names=True)
print(t2)
#    /-
# --|
#   |   /-B
#    \-|
#       \-H

Answer 1

[Kevin]

Hi Andrzej,

I was looking for something similar and ended up developing the following code.

def name_internal_nodes(T):
    for node in T.traverse():
        if node.is_leaf()==False:
            # list names of leaves
            leaf_names=[leaf.name for leaf in node.iter_leaves()]
            names_unique = list(set(leaf_names))
            # if all leaves have the same name, give that name to the node
            if (len(names_unique)==1):
                node.name = names_unique[0]

Basically, this function adds a name to an internal node if all children leaves of that node have the same name. Then, collapsed_leaf will keep the internal name with its support value.

On your example, it gives the following:

t1 = Tree('(("A. thaliana":0.723274,"A. thaliana":0.567784):0.067192,("B":0.279326,"H":0.756049):0.807788);', quoted_node_names=True)
print(t1.get_ascii(show_internal=True))
#      /-A. thaliana
#   /-|
#  |   \-A. thaliana
#--|
#  |   /-B
#   \-|
#      \-H
name_internal_nodes(t1)
print(t1.get_ascii(show_internal=True))
#              /-A. thaliana
#   /A. thaliana
#  |           \-A. thaliana
#--|
#  |   /-B
#   \-|
#      \-H
# Cache node content
node2labels = t1.get_cached_content(store_attr="name")
# Collapse tree
t2 = Tree( t1.write(is_leaf_fn=collapsed_leaf, quoted_node_names=True), quoted_node_names=True)
print(t2.get_ascii(show_internal=True))
#   /-A. thaliana
#--|
#  |   /-B
#   \-|
#      \-H
print (t2.write())
#(A. thaliana:0.067192,(B:0.279326,H:0.756049)1:0.807788);

[Andrzej Zielezinski]

Thank you, Kevin. This is exactly what I was looking for. You saved my life :)

[susan.klein]

Hi, I have tried to adapt this to collapse a tree based on the first string in leaf labels (being a genus name), but it is not being collapsed or renamed. Can you suggest what the problem might be?

Thanks for any help.

test.tree = (("Arabidopsis_thaliana1":0.723274,"Arabidopsis_thaliana2":0.567784):0.067192,("B":0.279326,"H":0.756049):0.807788);

import ete3
import sys

def name_internal_nodes(T):

    for node in T.traverse():
        if node.is_leaf()==False:
            # list names of leaves
            leaf_names=[leaf.name.split("_")[0] for leaf in node.iter_leaves()]

            names_unique = list(set(leaf_names))
            print(names_unique)
            # if all leaves have the same name, give that name to the node
            if (len(names_unique)==1):
                node.name = names_unique[0]

    return T

def collapsed_leaf(node):
    if len(node2labels[node]) == 1:
        return True
    else:
        return False

#global
t1 = ete3.Tree(sys.argv[1], quoted_node_names=True)
node2labels = t1.get_cached_content(store_attr="name")
g=open(sys.argv[2],'w')
print(t1.get_ascii(show_internal=True))
t2=name_internal_nodes(t1)
print(t2.get_ascii(show_internal=True))
node2labels = t2.get_cached_content(store_attr="name")
t3 = ete3.Tree( t2.write(is_leaf_fn=collapsed_leaf, quoted_node_names=True), quoted_node_names=True)
print(t3.get_ascii(show_internal=True))
g.write(str(t3.write()))

[susan.klein]

Ok, think I solved it - had to also rename the leaves:

#!/usr/bin/env python3


import ete3
import sys

def name_internal_nodes(T):

    for node in T.traverse():
        if node.is_leaf()==False:
            # list names of leaves
            leaf_names=[leaf.name.split("_")[0] for leaf in node.iter_leaves()]

            c=-1
            for leaf in node.iter_leaves():
                c=c+1
                leaf.name=leaf_names[c]

            names_unique = list(set(leaf_names))
            print(names_unique)
            # if all leaves have the same name, give that name to the node
            if (len(names_unique)==1):
                node.name = names_unique[0]

    return T

def collapsed_leaf(node):
    if len(node2labels[node]) == 1:
        return True
    else:
        return False

#global
t1 = ete3.Tree(sys.argv[1], quoted_node_names=True)
node2labels = t1.get_cached_content(store_attr="name")
g=open(sys.argv[2],'w')
print(t1.get_ascii(show_internal=True))
t2=name_internal_nodes(t1)
print(t2.get_ascii(show_internal=True))
node2labels = t2.get_cached_content(store_attr="name")
t3 = ete3.Tree(t2.write(is_leaf_fn=collapsed_leaf, quoted_node_names=True), quoted_node_names=True)
print(t3.get_ascii(show_internal=True))
g.write(str(t3.write()))