def test_no_similarity(self):
filename_a = os.path.join("data", "276.pdb")
filename_b = os.path.join("data", "4629.pdb")
activesite_a = io.read_active_site(filename_a)
activesite_b = io.read_active_site(filename_b)
seq_a = cluster.convert_to_aa_str(activesite_a.residues)
seq_b = cluster.convert_to_aa_str(activesite_b.residues)
# should be 8 because there are 8 edit operations between the strings
self.assertEqual(cluster.compute_similarity(seq_a, seq_b), 8)
def test_identical_similarity(self):
filename_a = os.path.join("data", "276.pdb")
filename_b = os.path.join("data", "276.pdb")
activesite_a = io.read_active_site(filename_a)
activesite_b = io.read_active_site(filename_b)
seq_a = cluster.convert_to_aa_str(activesite_a.residues)
seq_b = cluster.convert_to_aa_str(activesite_b.residues)
# should be 0 because the strings are identical
self.assertEqual(cluster.compute_similarity(seq_a, seq_b), 0)
def test_residues(filename, names, numbers):
filepath = os.path.join("data", filename)
activesite = io.read_active_site(filepath)
assert [residue.type for residue in activesite.residues] == names
assert [residue.number for residue in activesite.residues] == numbers
def test_atoms(filename, residue_number, atoms, xs, ys, zs):
filepath = os.path.join("data", filename)
activesite = io.read_active_site(filepath)
residue = activesite.residues[residue_number]
assert [atom.type for atom in residue.atoms] == atoms
assert [atom.coords for atom in residue.atoms] == list(zip(xs, ys, zs))
def test_partition_clustering_one(self):
# tractable subset
pdb_ids = [276, 4629, 10701]
active_sites = []
for id in pdb_ids:
filepath = os.path.join("data", "%i.pdb" % id)
active_sites.append(io.read_active_site(filepath))
clusters = cluster.cluster_by_partitioning(active_sites, 1)[0]
# Should all be in one cluster
self.assertClustersEqual(clusters, [["276", "10701", "4629"]])
def test_partition_clustering_individual(self):
# tractable subset
pdb_ids = [276, 4629, 10701]
active_sites = []
for id in pdb_ids:
filepath = os.path.join("data", "%i.pdb" % id)
active_sites.append(io.read_active_site(filepath))
clusters = cluster.cluster_by_partitioning(active_sites,
len(pdb_ids))[0]
# Should each be in their own cluster.
self.assertClustersEqual(clusters, [["276"], ["10701"], ["4629"]])
def test_hierarchical_clustering_two(self):
# tractable subset
pdb_ids = [37438, 55996, 81859]
# Try it four times, since test is flaky (because of inherent randomness)
active_sites = []
for id in pdb_ids:
filepath = os.path.join("data", "%i.pdb" % id)
active_sites.append(io.read_active_site(filepath))
clusters = cluster.cluster_hierarchically(active_sites, 2)[0]
# 37438 should be clustered with 81859
self.assertClustersEqual(clusters, [["37438", "55996"], ["81859"]])
def test_partition_clustering_two(self):
# tractable subset
pdb_ids = [37438, 55996, 81859]
# Try it four times, since test is flaky (because of inherent randomness)
# Normally would use flaky.
for i in range(4):
try:
active_sites = []
for id in pdb_ids:
filepath = os.path.join("data", "%i.pdb" % id)
active_sites.append(io.read_active_site(filepath))
clusters = cluster.cluster_by_partitioning(active_sites, 2)[0]
# 37438 should be clustered with 81859
self.assertClustersEqual(clusters,
[["37438", "55996"], ["81859"]])
break
except AssertionError:
pass
else:
raise AssertionError("Partitioning failed on all 4 attempts")