def test_neighbor_joining():
"""
Compare the results of `neighbor_join()` with a known tree.
"""
dist = np.array([
[0, 5, 4, 7, 6, 8],
[5, 0, 7, 10, 9, 11],
[4, 7, 0, 7, 6, 8],
[7, 10, 7, 0, 5, 9],
[6, 9, 6, 5, 0, 8],
[8, 11, 8, 9, 8, 0],
])
ref_tree = phylo.Tree(
phylo.TreeNode([
phylo.TreeNode([
phylo.TreeNode([
phylo.TreeNode(index=0),
phylo.TreeNode(index=1),
], [1, 4]),
phylo.TreeNode(index=2),
], [1, 2]),
phylo.TreeNode([
phylo.TreeNode(index=3),
phylo.TreeNode(index=4),
], [3, 2]),
phylo.TreeNode(index=5),
], [1, 1, 5]))
test_tree = phylo.neighbor_joining(dist)
assert test_tree == ref_tree
fasta_file = fasta.FastaFile.read(
entrez.fetch_single_file(ids,
file_name=None,
db_name="protein",
ret_type="fasta"))
sequences = [seq.ProteinSequence(seq_str) for seq_str in fasta_file.values()]
# Create multiple sequence alignment with Clustal Omega
alignment = clustalo.ClustalOmegaApp.align(sequences)
# The distance measure required for the tree calculation is the
# percentage of non-identical amino acids in the respective two
# sequences
distances = 1 - align.get_pairwise_sequence_identity(alignment,
mode="shortest")
# Create tree via neighbor joining
tree = phylo.neighbor_joining(distances)
# Convert to NetworkX graph
#For the graph visualization, the edge directions are unnecessary
graph = tree.as_graph().to_undirected()
fig = plt.figure(figsize=(8.0, 8.0))
ax = fig.gca()
ax.axis("off")
# Calculate position of nodes in the plot
pos = nx.kamada_kawai_layout(graph)
# Assign the gene names to the nodes that represent a reference index
node_labels = {i: name for i, name in enumerate(genes)}
nx.draw_networkx(
graph,
pos,
ax=ax,