def test_binaries_zero_cutoff():
trajectories = MinimalFsPeptide().get_cached().trajectories
binarycontactfeaturizer = BinaryContactFeaturizer(cutoff=0)
binaries = binarycontactfeaturizer.transform(trajectories)
assert binaries[0].shape[1] == 171
assert np.sum(binaries[0]) == 0
def test_binaries_zero_cutoff():
trajectories = MinimalFsPeptide().get_cached().trajectories
binarycontactfeaturizer = BinaryContactFeaturizer(cutoff=0)
binaries = binarycontactfeaturizer.transform(trajectories)
assert binaries[0].shape[1] == 171
assert np.sum(binaries[0]) == 0
def test_binary_to_logistics():
trajectories = MinimalFsPeptide().get_cached().trajectories
steepness = np.absolute(10 * np.random.randn())
center = np.absolute(np.random.randn())
binarycontactfeaturizer = BinaryContactFeaturizer(cutoff=center)
binaries = binarycontactfeaturizer.transform(trajectories)
logisticcontactfeaturizer = LogisticContactFeaturizer(center=center, steepness=steepness)
logistics = logisticcontactfeaturizer.transform(trajectories)
# This checks that no distances that are larger than the center are logistically
# transformed such that they are less than 1/2
np.testing.assert_array_almost_equal(binaries[0], logistics[0] > 0.5)
def test_binary_to_logistics():
trajectories = MinimalFsPeptide().get_cached().trajectories
steepness = np.absolute(10 * np.random.randn())
center = np.absolute(np.random.randn())
binarycontactfeaturizer = BinaryContactFeaturizer(cutoff=center)
binaries = binarycontactfeaturizer.transform(trajectories)
logisticcontactfeaturizer = LogisticContactFeaturizer(center=center,
steepness=steepness)
logistics = logisticcontactfeaturizer.transform(trajectories)
# This checks that no distances that are larger than the center are logistically
# transformed such that they are less than 1/2
np.testing.assert_array_almost_equal(binaries[0], logistics[0] > 0.5)
def create_equivalent_contact_featurizer(yaml_file, alignment_file,
protein_list=None,
pairs=None,
same_residue=True,
transform=None,
**kwargs):
"""
Create a equivalent contacts featurizer for a set of proteins
:param yaml_file: yaml file location
:param alignment_file: alignment file location
:param pairs: wanted sequence index positions in the alignment
You need to just figure out the wanted location for one residue.
_map_residue_ind_seq_ind function can help with this
:same residue: True is you would restrict to having the same residue at the same
sequence position.
:param kwargs: kwargs for the contact featurizer
:return: dictionary of contact featurizers. one for each protein
"""
featurizer_dict={}
#load alignment file
yaml_file = load_yaml_file(yaml_file)
alignment_file = _parse_alignment_file(alignment_file)
if protein_list is None:
protein_list = yaml_file["protein_list"]
if pairs is None:
#use the max length(probably a horrible idea)
max_seq_len = max([len(alignment_file[i]) for i in alignment_file.keys()])
pairs = [i for i in itertools.combinations(range(max_seq_len), 2)]
for protein in protein_list:
print(protein)
#get a list of residues we can keep
can_keep=[]
#get mapping and seq
prt_mapping, prt_seq = _map_residue_ind_seq_ind(yaml_file, protein,
alignment_file[protein])
#for wanted positions in the massive wanted indices list
inv_map = {v: k for k, v in prt_mapping.items()}
for position in np.unique(pairs):
#get the
#get the possible codes at every position
possible_codes = set([alignment_file[p][position] for p in alignment_file.keys()])
#if there is not a missing residue
if not "-" in possible_codes:
if same_residue and len(set(possible_codes))!=1:
continue
# get the inverse mapping and add it to the list of can keep
residue_index = inv_map[position]
can_keep.append(residue_index)
#sort it because i dont want random bs issues.
can_keep = np.sort(can_keep)
#get its pairs
actual_pairs = np.array([i for i in itertools.combinations(can_keep, 2) if i in pairs])
if transform=='logistic':
featurizer_dict[protein] = LogisticContactFeaturizer(contacts=actual_pairs, **kwargs)
elif transform=='binary':
featurizer_dict[protein] = BinaryContactFeaturizer(contacts=actual_pairs, **kwargs)
elif transform is None or transform=="none":
featurizer_dict[protein] = ContactFeaturizer(contacts=actual_pairs, **kwargs)
else:
raise ValueError("type needs to be one of logistic, binary, none")
return featurizer_dict
