def fit(self, inputs, vectorizer):
"""
Parameters
----------
inputs: sequence list
vectorizer: a vectorizer
Returns
-------
self
"""
self.vectorizer = vectorizer
self.NNmodel = EdenNNF(n_neighbors=4)
self.NNmodel.fit(inputs)
return self
class GraphTransformerForgi(GraphTransformer):
def __init__(self):
'''
Parameters
----------
base_thickness_list: list of int
thickness list for the base graph
structure_mod : bool
should we introduce "F" nodes to keep multiloop flexible regarding substitution
include_base : base
if asked for all cips, i will also yield "normal" cips (whose core is not radius of abstract, but radius of base graph)
ignore_inserts:
bolges will be ignored and merged to their adjacend stems
Returns
-------
'''
def fit(self, inputs, vectorizer):
"""
Parameters
----------
inputs: sequence list
vectorizer: a vectorizer
Returns
-------
self
"""
self.vectorizer = vectorizer
self.NNmodel = EdenNNF(n_neighbors=4)
self.NNmodel.fit(inputs)
return self
def fit_transform(self, inputs):
"""
Parameters
----------
inputs: sequences
Returns
-------
many graphdecomposers
"""
inputs = list(inputs)
self.fit(inputs, self.vectorizer)
inputs = [b for a, b in inputs]
return self.transform(inputs)
def re_transform_single(self, graph):
"""
Parameters
----------
graph: digraph
Returns
-------
graph decomposer
"""
try:
sequence = get_sequence(graph)
except:
logger.debug('sequenceproblem: this is not an rna')
# draw.graphlearn(graph, size=20)
return None
sequence = sequence.replace("F", '')
trans = self.transform([sequence])[0]
# if trans._base_graph.graph['energy'] > -10:
# return None
return trans
def abstract_graph(self):
'''
we need to make an abstraction Ooo
'''
# create the abstract graph and populate the contracted set
abstract_graph = forgi.get_abstr_graph(self.structure, ignore_inserts=self.ignore_inserts)
abstract_graph = self.vectorizer._edge_to_vertex_transform(abstract_graph)
completed_abstract_graph = forgi.edge_parent_finder(abstract_graph, self._base_graph)
# eden is forcing us to set a label and a contracted attribute.. lets do this
for n, d in completed_abstract_graph.nodes(data=True):
if 'edge' in d:
d['label'] = 'e'
# in the abstract graph , all the edge nodes need to have a contracted attribute.
# originaly this happens naturally but since we make multiloops into one loop
# there are some left out
for n, d in completed_abstract_graph.nodes(data=True):
if 'contracted' not in d:
d['contracted'] = set()
return completed_abstract_graph
def transform(self, sequences):
"""
Parameters
----------
sequences : iterable over rna sequences
Returns
-------
list of RnaGraphWrappers
"""
result = []
for sequence in sequences:
# if we eat a tupple, it musst be a (name, sequence) type :) we only want a sequence
if type(sequence) == type(()):
logger.warning('YOUR INPUT IS A TUPPLE, GIVE ME A SEQUENCE, SINCERELY -- YOUR RNA PREPROCESSOR')
# get structure
structure, energy, sequence = self.NNmodel.transform_single(('fake', sequence))
# FIXING STRUCTURE
structure, sequence = fix_structure(structure, sequence)
if structure == None:
result.append(None)
continue
# built base_graph
base_graph = converter.sequence_dotbracket_to_graph(seq_info=sequence, seq_struct=structure)
base_graph = self.vectorizer._edge_to_vertex_transform(base_graph)
base_graph = expanded_rna_graph_to_digraph(base_graph)
base_graph.graph['energy'] = energy
base_graph.graph['sequence'] = sequence
base_graph.graph['structure'] = structure
result.append(
(sequence, structure, base_graph, self.abstract_graph(base_graph))
)
return result
