This module provides functionality for working with trees, including phylogenetic trees and hierarchies, and prefix trees (i.e., tries). Functionality is provided for constructing trees, for traversing in multiple ways, comparisons, fetching subtrees, and more. This module supports trees that are multifurcating and nodes that have single descendants.
TreeNode([name, length, parent, children])  Representation of a node within a tree 
CompressedTrie([pair_list])  A compressed Trie for a list of (key, value) pairs 
nj(dm[, disallow_negative_branch_length, ...])  Apply neighbor joining for phylogenetic reconstruction. 
fasta_to_pairlist(seqs)  Yields (key, value) pairs, useful for populating a Trie object 
>>> from skbio.core.tree import TreeNode
A new tree can be constructed from a Newick string. Newick is a common format used to represent tree objects within a file. Newick was part of the original PHYLIP package from Joseph Felsenstein’s group (defined here), and is based around representing nesting with parentheses. For instance, the following string describes a 3 taxon tree, with one internal node:
((A, B)C, D)root;
Where A, B, and D are tips of the tree, and C is an internal node that covers tips A and B.
Now let’s construct a simple tree and dump an ASCII representation:
>>> tree = TreeNode.from_newick("((A, B)C, D)root;")
>>> print tree.is_root() # is this the root of the tree?
True
>>> print tree.is_tip() # is this node a tip?
False
>>> print tree.ascii_art()
/A
/C
root \B

\D
There are a few common ways to traverse a tree, and depending on your use, some methods are more appropriate than others. Wikipedia has a well written page on tree traversal methods, and will go into further depth than what we’ll cover here. We’re only going to cover two of the commonly used traversals here, preorder and postorder, but we will show examples of two other common helper traversal methods to gather tips or internal nodes.
The first traversal we’ll cover is a preorder traversal in which you evaluate from root to tips, looking at the left most child first. For instance:
>>> for node in tree.preorder():
... print node.name
root
C
A
B
D
The next method we’ll look at is a postorder traveral which will evaluate the left subtree tips first before walking back up the tree:
>>> for node in tree.postorder():
... print node.name
A
B
C
D
root
TreeNode provides two helper methods as well for iterating over just the tips or for iterating over just the internal nodes.
>>> for node in tree.tips():
... print "Node name: %s, Is a tip: %s" % (node.name, node.is_tip())
Node name: A, Is a tip: True
Node name: B, Is a tip: True
Node name: D, Is a tip: True
>>> for node in tree.non_tips():
... print "Node name: %s, Is a tip: %s" % (node.name, node.is_tip())
Node name: C, Is a tip: False
Note, by default, non_tips will ignore self (which is the root in this case). You can pass the include_self flag to non_tips if you wish to include self.
The TreeNode provides a few ways to compare trees. First, let’s create two similar trees and compare their topologies using compare_subsets. This distance is the fraction of common clades present in the two trees, where a distance of 0 means the trees contain identical clades, and a distance of 1 indicates the trees do not share any common clades:
>>> tree1 = TreeNode.from_newick("((A, B)C, (D, E)F, (G, H)I)root;")
>>> tree2 = TreeNode.from_newick("((G, H)C, (D, E)F, (B, A)I)root;")
>>> tree3 = TreeNode.from_newick("((D, B)C, (A, E)F, (G, H)I)root;")
>>> print tree1.compare_subsets(tree1) # identity case
0.0
>>> print tree1.compare_subsets(tree2) # same tree but different clade order
0.0
>>> print tree1.compare_subsets(tree3) # only 1 of 3 common subsets
0.666666666667
We can additionally take into account branch length when computing distances between trees. First, we’re going to construct two new trees with described branch length, note the difference in the Newick strings:
>>> tree1 = TreeNode.from_newick("((A:0.1, B:0.2)C:0.3, D:0.4, E:0.5)root;")
>>> tree2 = TreeNode.from_newick("((A:0.4, B:0.8)C:0.3, D:0.1, E:0.5)root;")
In these two trees, we’ve added on a description of length from the node to its parent, so for instance:
>>> for node in tree1.postorder():
... print node.name, node.length
A 0.1
B 0.2
C 0.3
D 0.4
E 0.5
root None
Now let’s compare two trees using the distances computed pairwise between tips in the trees. The distance computed, by default, is the correlation of all pairwise tiptotip distances between trees:
>>> print tree1.compare_tip_distances(tree1) # identity case
0.0
>>> print tree1.compare_tip_distances(tree2)
0.120492524415
Construct a Trie from a (key, value) list
>>> from skbio.core.tree import CompressedTrie
>>> pair_list = [("ab", "0"),
... ("abababa", "1"),
... ("abab", "2"),
... ("baba", "3"),
... ("ababaa", "4"),
... ("a", "5"),
... ("abababa", "6"),
... ("bab", "7"),
... ("babba", "8")]
>>> t = CompressedTrie(pair_list)
Get the number of keys stored in the trie
>>> len(t)
9
Get the number of nodes in the trie
>>> t.size
10
Get the trie’s prefix map
>>> t.prefix_map
{'1': ['6', '2', '0', '5'], '8': ['7'], '3': [], '4': []}
Find the value attached to a given key
>>> t.find("ababaa")
['4']
Add a new (key, value) pair to the Trie
>>> t.insert("bac", "9")
>>> t.find("bac")
['9']
>>> t.prefix_map
{'1': ['6', '2', '0', '5'], '9': [], '3': [], '4': [], '8': ['7']}
Create a new trie with a list of sequences
>>> from skbio.core.tree import fasta_to_pairlist
>>> seqs = [("s0", "ACA"),
... ("s1", "ACAGTC"),
... ("s2", "ACTA"),
... ("s3", "CAGT"),
... ("s4", "CATGAA"),
... ("s5", "A"),
... ("s6", "CATGTA"),
... ("s7", "CACCA")]
>>> t = CompressedTrie(fasta_to_pairlist(seqs))
>>> t.prefix_map
{'s3': [], 's2': [], 's1': ['s0', 's5'], 's7': [], 's6': [], 's4': []}