def ubiquitize_model(in_sbml,
chebi,
groups_sbml,
ub_s_ids=None,
ub_chebi_ids=None):
"""
Infers and marks ubiquitous species in the model.
:param in_sbml: str, path to the input SBML file
:param chebi: mod_sbml.onto.obo_ontology.Ontology ChEBI ontology
:param groups_sbml: str, path to the output SBML file (with groups extension)
:param ub_s_ids: optional, ids of ubiquitous species (will be inferred if set to None)
:param ub_chebi_ids: optional, ids of ubiquitous ChEBI terms (will be inferred if set to None)
:return: tuple (s_id2chebi_id, ub_s_ids): dict {species_id: ChEBI_term_id}, collection of ubiquitous species_ids.
"""
input_doc = libsbml.SBMLReader().readSBML(in_sbml)
input_model = input_doc.getModel()
annotate_metabolites(input_model, chebi)
logging.info("mapping species to ChEBI")
s_id2chebi_id = get_species_id2chebi_id(input_model)
_, ub_s_ids = get_ub_elements(input_model, chebi, s_id2chebi_id,
ub_chebi_ids, ub_s_ids)
save_as_comp_generalized_sbml(input_model, None, groups_sbml, {}, {},
ub_s_ids, chebi)
return s_id2chebi_id, ub_s_ids
def update_model_element_ids(m_id, model, go2c_id, go, chebi):
id2id = {}
if need_boundary_compartment(model):
separate_boundary_metabolites(model)
annotate_metabolites(model, chebi)
annotate_compartments(model, go)
for c in model.getListOfCompartments():
c_id = c.getId()
go_id = get_go_id(c)
if not go_id:
go_id = c.getName()
if go_id:
go_id = go_id.lower()
if go_id == CYTOSOL and (CYTOSOL not in go2c_id) and (CYTOPLASM
in go2c_id):
go_id = CYTOPLASM
elif go_id == CYTOPLASM and (CYTOSOL
in go2c_id) and (CYTOPLASM
not in go2c_id):
go_id = CYTOSOL
if go_id not in go2c_id:
go2c_id[go_id] = "c_%s__%s" % (m_id, c_id)
new_c_id = go2c_id[go_id]
else:
new_c_id = "c_%s__%s" % (m_id, c_id)
c.setId(new_c_id)
id2id[c_id] = new_c_id
for c in model.getListOfCompartments():
if c.getOutside():
c.setOutside(id2id[c.getOutside()])
for s in model.getListOfSpecies():
if s.getCompartment():
s.setCompartment(id2id[s.getCompartment()])
old_id = s.getId()
new_id = "s_%s__%s" % (m_id, old_id)
s.setId(new_id)
id2id[old_id] = new_id
if s.getSpeciesType():
s.unsetSpeciesType()
for r in model.getListOfReactions():
old_id = r.getId()
new_id = "r_%s__%s" % (m_id, old_id)
r.setId(new_id)
id2id[old_id] = new_id
if r.getCompartment():
r.setCompartment(id2id[r.getCompartment()])
for s_ref in r.getListOfReactants():
s_ref.setSpecies(id2id[s_ref.getSpecies()])
for s_ref in r.getListOfProducts():
s_ref.setSpecies(id2id[s_ref.getSpecies()])
for s_ref in r.getListOfModifiers():
s_ref.setSpecies(id2id[s_ref.getSpecies()])
def add_boundary_metabolites(in_sbml, out_sbml):
"""
Creates a boundary compartment with the id 'Boundary',
and transforms each input/output reaction '*_e <-> ' into '*_e <-> *_b'.
:param in_sbml: path to the SBML file with the original model
:param out_sbml: path where to store the resulting SBML file
"""
input_doc = libsbml.SBMLReader().readSBML(in_sbml)
model = input_doc.getModel()
chebi = parse_simple(get_chebi())
annotate_metabolites(model, chebi)
create_boundary_species_in_boundary_reactions(model)
libsbml.SBMLWriter().writeSBMLToFile(input_doc, out_sbml)
def generalize_model(in_sbml, chebi, groups_sbml, out_sbml, ub_s_ids=None, ub_chebi_ids=None, ignore_biomass=True):
"""
Generalizes a model.
:param in_sbml: str, path to the input SBML file
:param chebi: mod_sbml.onto.obo_ontology.Ontology ChEBI ontology
:param groups_sbml: str, path to the output SBML file (with groups extension)
:param out_sbml: str, path to the output SBML file (generalized)
:param ub_s_ids: optional, ids of ubiquitous species (will be inferred if set to None)
:param ub_chebi_ids: optional, ids of ubiquitous ChEBI terms (will be inferred if set to None)
:param ignore_biomass: boolean, whether to ignore the biomass reaction (and its stoichiometry preserving constraint)
:return: tuple (r_id2g_eq, s_id2gr_id, s_id2chebi_id, ub_s_ids):
dict {reaction_id: reaction_group_id}, dict {species_id: species_group_id}, dict {species_id: ChEBI_term_id},
collection of ubiquitous species_ids.
"""
# input_model
input_doc = libsbml.SBMLReader().readSBML(in_sbml)
input_model = input_doc.getModel()
r_ids_to_ignore = get_biomass_r_ids(input_model) if ignore_biomass else None
remove_is_a_reactions(input_model)
annotate_metabolites(input_model, chebi)
# TODO: fix comp separation
# separate_boundary_metabolites(input_model)
remove_unused_elements(input_model)
logging.info("mapping species to ChEBI")
s_id2chebi_id = get_species_id2chebi_id(input_model)
ub_chebi_ids, ub_s_ids = get_ub_elements(input_model, chebi, s_id2chebi_id, ub_chebi_ids, ub_s_ids)
terms = (t for t in (chebi.get_term(t_id) for t_id in s_id2chebi_id.values()) if t)
old_onto_len = len(chebi)
filter_ontology(chebi, terms, relationships=EQUIVALENT_RELATIONSHIPS, min_deepness=3)
logging.info('Filtered the ontology from %d terms to %d' % (old_onto_len, len(chebi)))
threshold = min(max(3, int(0.1 * input_model.getNumReactions())), UBIQUITOUS_THRESHOLD)
s_id2clu, ub_s_ids = generalize_species(input_model, s_id2chebi_id, ub_s_ids, chebi, ub_chebi_ids, threshold,
r_ids_to_ignore=r_ids_to_ignore)
logging.info("generalized species")
r_id2clu = generalize_reactions(input_model, s_id2clu, s_id2chebi_id, ub_chebi_ids,
r_ids_to_ignore=r_ids_to_ignore)
logging.info("generalized reactions")
clu2s_ids = {(c_id, term): s_ids for ((c_id, (term, )), s_ids) in invert_map(s_id2clu).items()}
r_id2g_eq, s_id2gr_id = save_as_comp_generalized_sbml(input_model, out_sbml, groups_sbml, r_id2clu, clu2s_ids,
ub_s_ids, chebi)
return r_id2g_eq, s_id2gr_id, s_id2chebi_id, ub_s_ids
def ubiquitize_model(in_sbml, chebi, groups_sbml, ub_s_ids=None, ub_chebi_ids=None):
"""
Infers and marks ubiquitous species in the model.
:param in_sbml: str, path to the input SBML file
:param chebi: mod_sbml.onto.obo_ontology.Ontology ChEBI ontology
:param groups_sbml: str, path to the output SBML file (with groups extension)
:param ub_s_ids: optional, ids of ubiquitous species (will be inferred if set to None)
:param ub_chebi_ids: optional, ids of ubiquitous ChEBI terms (will be inferred if set to None)
:return: tuple (s_id2chebi_id, ub_s_ids): dict {species_id: ChEBI_term_id}, collection of ubiquitous species_ids.
"""
input_doc = libsbml.SBMLReader().readSBML(in_sbml)
input_model = input_doc.getModel()
annotate_metabolites(input_model, chebi)
logging.info("mapping species to ChEBI")
s_id2chebi_id = get_species_id2chebi_id(input_model)
_, ub_s_ids = get_ub_elements(input_model, chebi, s_id2chebi_id, ub_chebi_ids, ub_s_ids)
save_as_comp_generalized_sbml(input_model, None, groups_sbml, {}, {}, ub_s_ids, chebi)
return s_id2chebi_id, ub_s_ids
def generalize_model(in_sbml,
chebi,
groups_sbml,
out_sbml,
ub_s_ids=None,
ub_chebi_ids=None,
ignore_biomass=True):
"""
Generalizes a model.
:param in_sbml: str, path to the input SBML file
:param chebi: mod_sbml.onto.obo_ontology.Ontology ChEBI ontology
:param groups_sbml: str, path to the output SBML file (with groups extension)
:param out_sbml: str, path to the output SBML file (generalized)
:param ub_s_ids: optional, ids of ubiquitous species (will be inferred if set to None)
:param ub_chebi_ids: optional, ids of ubiquitous ChEBI terms (will be inferred if set to None)
:param ignore_biomass: boolean, whether to ignore the biomass reaction (and its stoichiometry preserving constraint)
:return: tuple (r_id2g_eq, s_id2gr_id, s_id2chebi_id, ub_s_ids):
dict {reaction_id: reaction_group_id}, dict {species_id: species_group_id}, dict {species_id: ChEBI_term_id},
collection of ubiquitous species_ids.
"""
# input_model
input_doc = libsbml.SBMLReader().readSBML(in_sbml)
input_model = input_doc.getModel()
r_ids_to_ignore = get_biomass_r_ids(
input_model) if ignore_biomass else None
remove_is_a_reactions(input_model)
annotate_metabolites(input_model, chebi)
# TODO: fix comp separation
# separate_boundary_metabolites(input_model)
remove_unused_elements(input_model)
logging.info("mapping species to ChEBI")
s_id2chebi_id = get_species_id2chebi_id(input_model)
ub_chebi_ids, ub_s_ids = get_ub_elements(input_model, chebi, s_id2chebi_id,
ub_chebi_ids, ub_s_ids)
terms = (t for t in (chebi.get_term(t_id)
for t_id in s_id2chebi_id.values()) if t)
old_onto_len = len(chebi)
filter_ontology(chebi,
terms,
relationships=EQUIVALENT_RELATIONSHIPS,
min_deepness=3)
logging.info('Filtered the ontology from %d terms to %d' %
(old_onto_len, len(chebi)))
threshold = min(max(3, int(0.1 * input_model.getNumReactions())),
UBIQUITOUS_THRESHOLD)
s_id2clu, ub_s_ids = generalize_species(input_model,
s_id2chebi_id,
ub_s_ids,
chebi,
ub_chebi_ids,
threshold,
r_ids_to_ignore=r_ids_to_ignore)
logging.info("generalized species")
r_id2clu = generalize_reactions(input_model,
s_id2clu,
s_id2chebi_id,
ub_chebi_ids,
r_ids_to_ignore=r_ids_to_ignore)
logging.info("generalized reactions")
clu2s_ids = {(c_id, term): s_ids
for ((c_id, (term, )), s_ids) in invert_map(s_id2clu).items()}
r_id2g_eq, s_id2gr_id = save_as_comp_generalized_sbml(
input_model, out_sbml, groups_sbml, r_id2clu, clu2s_ids, ub_s_ids,
chebi)
return r_id2g_eq, s_id2gr_id, s_id2chebi_id, ub_s_ids
