def createSequence(modrefseq, filtered_ids=None, description=None):
emod = cbmpy.readSBML3FBC(modrefseq)
if len(emod.genes) == 0:
emod.createGeneAssociationsFromAnnotations()
geneseq = {}
if description == None:
secDescr = os.path.split(modrefseq)[-1]
else:
secDescr = description
if filtered_ids == None:
filtered_ids = emod.getReactionIds()
for r_ in emod.reactions:
if r_.getId() in filtered_ids:
for a in r_.annotation:
if a.startswith('gbank_seq_'):
newName = a.replace('gbank_seq_', '')
if newName not in geneseq:
if oldbiopython:
seq = Bio.Seq.Seq(
r_.getAnnotation(a), Bio.Alphabet.ProteinAlphabet()
)
else:
seq = Bio.Seq.Seq(r_.getAnnotation(a))
geneseq[newName] = Bio.SeqRecord.SeqRecord(
seq,
id=newName,
name=newName,
description=secDescr,
annotations={"molecule_type": "protein"},
)
# print('Adding {}'.format(a))
print('{} genseqs added'.format(len(geneseq)))
return geneseq
def roundTripModelV1(self, model):
modname = 'rtv1.xml'
cbmpy.writeSBML3FBCV2(model, modname, directory=self.cDir)
del model
model = cbmpy.readSBML3FBC(os.path.join(self.cDir, modname),
xoptions={'validate': True})
return model
def setup_class(klass):
"""This method is run once for each class before any tests are run"""
klass.m = {}
for m in DATA:
if m.startswith('BIGG2_'):
cmod = cbmpy.readSBML3FBC(os.path.join(MDIR, m),
xoptions={'validate': True})
klass.m[m] = cmod
def test_run_fba_cobra(self):
print('\n\nThis test fails under Linux.\n\n')
cmod = cbmpy.readCOBRASBML('L2CBR_iJR904.glc.xml', work_dir=MDIR,\
output_dir=self.CDIR, delete_intermediate=True)
cmod2 = cbmpy.readSBML3FBC('L3FBCV1_iJR904.glc.xml',
work_dir=MDIR,
xoptions={'validate': True})
fba1 = cbmpy.doFBA(cmod)
fba2 = cbmpy.doFBA(cmod2)
del cmod, cmod2
assert_almost_equal(fba1, fba2)
def createSeqplusModelMetaIdx(fmod, oid, oclass, metadraft_lib_model):
"""
Create a MetaDraft template model index file from a seqplus model file::
- *fmod* the seqplus model file e.g. 'EstherDB.xml'
- *oid* the model unique shortname e.g. 'edb'
- *class* the model category e.g. 'vu'
- *metadraft_lib_model* the target MetaDraft lib_model directory
'edb', 'vu', "<path>"
"""
import cbmpy
dmod = cbmpy.readSBML3FBC(fmod)
fgb = os.path.abspath(fmod)
input_path, fmod = os.path.split(fgb)
# dmod.createGeneAssociationsFromAnnotations()
oclass = oclass.replace('-', '')
oid = oid.replace('-', '')
oid = '{}-{}'.format(oclass, oid)
if fmod.endswith('.xml'):
fmod = fmod[:-4]
fmod = '({})-({}).seqplus'.format(oid, fmod)
print(fmod)
linkDict = {}
linkDict[oid] = {}
linkDict["__idx__"] = {}
LD = linkDict[oid]
LD['genbank_in'] = fgb
LD['sbml_in'] = fgb
LD['data_path'] = input_path
LD['gene2reaction'] = dmod.getAllProteinGeneAssociations()
for g_ in LD['gene2reaction']:
linkDict['__idx__'][g_] = oid
LD['reaction2gene'] = dmod.getAllGeneProteinAssociations()
LD['taxon_id'] = "unknown"
LD['sbml_out'] = os.path.join(metadraft_lib_model, "{}.xml".format(fmod))
LD['sbml_out_generic'] = os.path.join(metadraft_lib_model, "{}.xml".format(fmod))
LD['fasta_out'] = None
Fj = open(os.path.join(input_path, '{}-link.json'.format(fmod)), 'w')
json.dump(linkDict, Fj, indent=1, separators=(',', ': '))
Fj.close()
cbmpy.writeSBML3FBC(
dmod,
os.path.join(input_path, fmod + '.xml'),
add_cbmpy_annot=True,
add_cobra_annot=False,
add_groups=False,
)
def setup_class(klass):
"""This method is run once for each class before any tests are run"""
klass.m = {}
for m in DATA:
cmod = None
if m.startswith('L2CBR_'):
cmod = cbmpy.readCOBRASBML(m, work_dir=MDIR,\
output_dir=self.CDIR, delete_intermediate=True)
elif m.startswith('L2FBA_'):
cmod = cbmpy.readSBML2FBA(os.path.join(MDIR, m))
elif m.startswith('L3FBCV1_') or m.startswith('L3FBCV2_'):
cmod = cbmpy.readSBML3FBC(os.path.join(MDIR, m),
xoptions={'validate': True})
if cmod is not None:
klass.m[m] = cmod
klass.CDIR = cDir
import os, time, numpy
cDir = os.path.dirname(os.path.abspath(os.sys.argv[0]))
import cbmpy as cbm
model_name = 'iAF692.xml'
cmod = cbm.readSBML3FBC(os.path.join(cDir, 'models', model_name))
cmodS = cmod.clone()
cbm.doFBA(cmod)
fva, fvan = cbm.CBMultiCore.runMultiCoreFVA(cmod, procs=4)
cbm.CBWrite.writeFVAtoCSV(fva,
fvan,
model_name.replace('.xml', ''),
fbaObj=cmod)
cbm.writeModelToExcel97(cmod, model_name.replace('.xml', ''))
exch_reactions = cmod.getReactionIds('R_EX_')
dump = []
# reactions are defined back to front so fwd is "o" and and backward is "i"
for xr_ in exch_reactions:
dump.append(cbm.CBTools.splitSingleReversibleReaction(cmodS, xr_, xr_.replace('R_EX_', 'R_EXo_'),\
xr_.replace('R_EX_', 'R_EXi_')))
cbm.doFBA(cmodS)
fva, fvan = cbm.CBMultiCore.runMultiCoreFVA(cmodS, procs=4)
cbm.CBWrite.writeFVAtoCSV(fva,
fvan,
model_name.replace('.xml', '.split'),
fbaObj=cmodS)
cbm.writeSBML3FBC(cmodS, model_name.replace('.xml', '.split.xml'))
cbm.CBWrite.writeModelToExcel97(cmodS, model_name.replace('.xml', '.split'))
import cbmpy as cbm
from examples_and_results.helpers_ECM_calc import *
"""CONSTANTS"""
model_name = "iND750"
# For a bigger computation, try:
# input_file_name = "bacteroid_ECMinputAll.csv"
# Define directories for finding models
model_dir = os.path.join(os.getcwd(), "models")
model_path = os.path.join(model_dir, model_name + ".xml")
mod = cbm.readSBML3FBC(model_path)
#
pairs = cbm.CBTools.findDeadEndReactions(mod)
external_metabolites, external_reactions = list(zip(
*pairs)) if len(pairs) else (list(
zip(*cbm.CBTools.findDeadEndMetabolites(mod)))[0], [])
# External according to Urbanczik
ext_urbanczik = [
'ac', 'acald', 'ala__L', 'co2', 'csn', 'ergst', 'etoh', 'gam6p', 'glc__D',
'hdcea', 'ocdcea', 'ocdcya', 'so4', 'xylt', 'zymst', 'nh4', 'asp__L',
'ser__L', 'fum', 'gly', 'thr__L'
]
force_feed = ['ac']
ext_urbanczik_inds = [
ind for ind, metab in enumerate(external_metabolites)
if metab[2:-2] in ext_urbanczik
def test_load_L3FBCV1_iJR904(self):
cmod = cbmpy.readSBML3FBC(os.path.join(MDIR, 'L3FBCV1_iJR904.glc.xml'),
xoptions={'validate': True})
assert_not_equal(cmod, None)
def checkModelLocusTags(sbml, genbank, allow_gene_names=False):
"""
Checks the gene identifiers (assuming they are locus tags) against a genbank file of the same organism
- *sbml* the model SBML (*.xml) file
- *genbank* the associated GenBank (*.gbk) full file_s) (including CDS annotations and sequences)
- *allow_gene_names* allow gene names, non-unique as gene identifiers if locus tags are not present. USE WITH CAUTION!!!
"""
cmod = cbmpy.readSBML3FBC(sbml)
cntr = 0
if type(genbank) == str:
genbank = [genbank]
gprMap = {}
gprMapAnnot = {}
gnoprMap = {}
fileNum = 0
no_locus_tag = []
for G in genbank:
print('CheckModelLocusTags is processing: {}'.format(G))
for seq_record in SeqIO.parse(G, "genbank"):
# print(seq_record.id)
cntr += 1
if cntr > 1:
print("INFO: Multiple sequences encountered in file: {}".format(G))
# raise RuntimeError, "\nMutltiple sequences encountered in file: {}".format(G)
# print(repr(seq_record.seq))
# print(len(seq_record))
# add all the annotations from the genbank record(s) into the model? Use first record
if fileNum == 0:
gprMapAnnot[seq_record.id] = seq_record.annotations.copy()
try:
for r_ in gprMapAnnot[seq_record.id]['references']:
if r_.pubmed_id != '':
cmod.addMIRIAMannotation(
'isDescribedBy', 'PubMed', r_.pubmed_id
)
gprMapAnnot[seq_record.id].pop('references')
cmod.setAnnotation('genbank_id', seq_record.id)
cmod.setAnnotation('genbank_name', seq_record.name)
except KeyError:
print('checkModelLocusTags: no references')
# global features
features = [f.qualifiers for f in seq_record.features if f.type == 'source']
if len(features) > 0:
features = features[0]
for f_ in features:
if f_ == 'db_xref':
for ff_ in features[f_]:
if ff_.startswith('taxon:'):
cmod.setAnnotation(
'genbank_taxon_id', ff_.replace('taxon:', '')
)
break
cmod.setAnnotation('genbank_{}'.format(f_), features[f_])
for r_ in gprMapAnnot[seq_record.id]:
cmod.setAnnotation(
'genbank_{}'.format(r_), gprMapAnnot[seq_record.id][r_]
)
fileNum += 1
GBFile = open(G, 'r')
USING_GENE_NAME = False
for cds in Bio.SeqIO.InsdcIO.GenBankCdsFeatureIterator(GBFile):
if cds.seq != None:
if cds.name != '<unknown name>':
gprMap[cds.name] = cds
# this is dangerous as there is no defined 1:1 mapping
elif 'gene' in cds.annotations and allow_gene_names:
gprMap[cds.annotations['gene']] = cds
USING_GENE_NAME = True
no_locus_tag.append(cds.annotations['gene'])
else:
gnoprMap[cds.name] = cds
else:
gnoprMap[cds.name] = cds
GBFile.close()
oldLoTags = []
# for g_ in cmod.getGeneIds():
for g_ in cmod.getGeneLabels():
if g_ not in gprMap:
# print(g_)
if g_ != 'None':
oldLoTags.append(g_)
# cbmpy.CBTools.pprint.pprint(gprMap)
oldtags = [
gprMap[a].annotations['old_locus_tag'].split(' ') + [a]
for a in gprMap
if 'old_locus_tag' in gprMap[a].annotations
]
oldtags2 = [
gnoprMap[a].annotations['old_locus_tag'].split(' ') + [a]
for a in gnoprMap
if 'old_locus_tag' in gnoprMap[a].annotations
]
old2new = {}
old2new2 = {}
for x in oldtags:
for y in x:
if y != x[-1]:
old2new[y] = x[-1]
for x in oldtags2:
for y in x:
if y != x[-1]:
old2new2[y] = x[-1]
print('\n\nChecking locus tags\n===================')
updated = {}
unknown = []
noseq = []
good = []
F = open(sbml.replace('.xml', '.seqcheck.csv'), 'w')
# if cmod.__FBC_VERSION__ < 2:
# geneIDs = cmod.getGeneIds()
# else:
# geneIDs = cmod.getGeneLabels()
geneIDs = cmod.getGeneLabels()
for g_ in geneIDs:
if g_ not in gprMap:
if g_ in old2new:
print('old {} --> new {}'.format(g_, old2new[g_]))
updated[g_] = old2new[g_]
F.write('UPDATED,{},{}\n'.format(g_, old2new[g_]))
elif g_ in old2new2:
print('NoSEQ: old {} --> new {}'.format(g_, old2new2[g_]))
noseq.append((g_, old2new2[g_]))
F.write('NOSEQ,{},{}\n'.format(g_, old2new2[g_]))
else:
print('UNKNOWN gene: {}'.format(g_))
unknown.append(g_)
F.write('UNKNOWN,{}\n'.format(g_))
else:
good.append(g_)
F.close()
if USING_GENE_NAME:
print(
'\nINFO: model contained {} genes without locus tags. For these genes \
the /gene name was used instead. This gene name may not be unique, please check model!!'.format(
len(no_locus_tag)
)
)
# print(no_locus_tag)
# cbmpy.CBTools.storeObj(gprMap, sbml.replace('.xml', '.seqplus.dat'))
return good, updated, noseq, unknown, cmod, gprMap, gprMapAnnot, no_locus_tag
def setup_class(klass):
"""This method is run once for each class before any tests are run"""
klass.mcore = cbmpy.readSBML3FBC('cbmpy_test_core')
klass.sboterm1 = 'SBO:1234567'
klass.cDir = cDir