def readPublicDatasets( nwdata, c ):
# OK, NOW we dump in the public datasets
for pd in c['public_dicts'] :
# (done) make emili follow biogrid field conventions
# TODO ditto bioplex
if pd['infilename'] in os.listdir('.') :
pdsf = open(pd['infilename'])
elif os.path.isfile(pd['infilename']) :
pdsf = open(pd['infilename'])
else :
pdsf = open(c['publicDatadir'] + pd['infilename'])
temporaryds = I.dataSet( i_filter = c['iact_filter'] )
if pd.get('convert') == 'm2h' :
temporaryds.parse( pdsf, fd = I.fd_biogrid, m2h = True, qualify = pd.get('qualify',''),
directed = False, force_qualify = True )
elif pd.get('convert') == 'h2m' :
temporaryds.parse( pdsf, fd = I.fd_biogrid, h2m = True, qualify = pd.get('qualify',''),
directed = False, force_qualify = True )
else :
temporaryds.parse( pdsf, fd = I.fd_biogrid, qualify = pd.get('qualify',''),
directed = False, force_qualify = True )
print( 'saving public dataset' + pd['infilename'])
sio = StringIO()
temporaryds.save( sio, edges = { e for e in temporaryds.edges.values()
if e.weight >= pd.get('minweight',0) and e.totalscore >= pd.get('minscore',0) });
sio.seek(0)
print( 'reloading public dataset' + pd['infilename'])
nwdata.load_from( sio )
sio.close()
pdsf.close()
def load_biogrid(filters='default',m2h=False,h2m=False,force_qualify=True,superdebug=False,qualify='') :
global biogrid ;
bgpath= REFERENCEPATH + 'biogrid_latest' ;
from lib import interactors as I
from lib import interactors_extras as ie
mynf=None ;
myif=None ;
(myif,mynf)=ie.bg_regex_assembler() ;
if filters == 'default' :
pass ;
elif filters.lower() in 'exogenous' :
mynf=ie.exogenous_regex_assembler() ;
elif filters.lower() in 'ionly' :
mynf=None ;
elif filters.lower() in 'physical' :
physical_subf=ie.subfilter(r'genetic','systemType') ;
myif=ie.Filter(excludes={physical_subf,}) ;
mynf=None ;
elif filters is None or filters.lower() in 'none' :
mynf=None ;
myif=None ;
biogrid=I.dataSet(i_filter=myif,n_filter=mynf,superdebug=superdebug) ;
bgfile=open(bgpath) ;
biogrid.parse(bgfile,fd=I.fd_biogrid,m2h=m2h,h2m=h2m,force_qualify=force_qualify,qualify=qualify) ;
bgfile.close()
def load_preppi() :
global preppi ;
from lib import interactors as I
from lib import interactors_extras as ie
preppipath= REFERENCEPATH + 'preppi_150727_lr600.i' ;
preppi=I.dataSet() ;
preppi_f=open(preppipath) ;
preppi.parse(preppi_f,fd=I.fdms) ;
preppi_f.close() ;
def load_preppi():
global preppi
from lib import interactors as I
from lib import interactors_extras as ie
preppipath = REFERENCEPATH + 'preppi_150727_lr600.i'
preppi = I.dataSet()
preppi_f = open(preppipath)
preppi.parse(preppi_f, fd=I.fdms)
preppi_f.close()
def readPublicDatasets(nwdata, c):
# OK, NOW we dump in the public datasets
for pd in c['public_dicts']:
# (done) make emili follow biogrid field conventions
# TODO ditto bioplex
if pd['infilename'] in os.listdir('.'):
pdsf = open(pd['infilename'])
elif os.path.isfile(pd['infilename']):
pdsf = open(pd['infilename'])
else:
pdsf = open(c['publicDatadir'] + pd['infilename'])
temporaryds = I.dataSet(i_filter=c['iact_filter'])
if pd.get('convert') == 'm2h':
temporaryds.parse(pdsf,
fd=I.fd_biogrid,
m2h=True,
qualify=pd.get('qualify', ''),
directed=False,
force_qualify=True)
elif pd.get('convert') == 'h2m':
temporaryds.parse(pdsf,
fd=I.fd_biogrid,
h2m=True,
qualify=pd.get('qualify', ''),
directed=False,
force_qualify=True)
else:
temporaryds.parse(pdsf,
fd=I.fd_biogrid,
qualify=pd.get('qualify', ''),
directed=False,
force_qualify=True)
print('saving public dataset' + pd['infilename'])
sio = StringIO()
temporaryds.save(sio,
edges={
e
for e in temporaryds.edges.values()
if e.weight >= pd.get('minweight', 0)
and e.totalscore >= pd.get('minscore', 0)
})
sio.seek(0)
print('reloading public dataset' + pd['infilename'])
nwdata.load_from(sio)
sio.close()
pdsf.close()
def load_refsuite(filters='default',
m2h=False,
h2m=False,
superdebug=False,
force=False):
from lib import interactors as I
from lib import interactors_extras as ie
global refsuite
if refsuite is not None and len(refsuite.nodes) > 0 and not force:
return
bgpath = REFERENCEPATH + 'biogrid_latest'
empath = REFERENCEPATH + '/complexes/emiliome.i'
bppath = REFERENCEPATH + '/bioplex.i'
mynf = None
myif = None
(myif, mynf) = ie.bg_regex_assembler()
if filters == 'default':
pass
elif filters.lower() in 'exogenous':
mynf = ie.exogenous_regex_assembler()
elif filters.lower() in 'ionly':
mynf = None
elif filters is None or filters.lower() in 'none':
mynf = None
myif = None
refsuite = I.dataSet(i_filter=myif, n_filter=mynf, superdebug=superdebug)
bgfile = open(bgpath)
refsuite.parse(bgfile,fd=I.fd_biogrid,m2h=m2h,h2m=h2m,force_qualify=True,\
force_score=0.0,qualify='bg')
bgfile.close()
emfile = open(empath)
refsuite.parse(emfile,fd=I.fd_emili,m2h=m2h,h2m=h2m,force_qualify=True,\
force_score=0.0,qualify='em')
emfile.close()
bpfile = open(bppath)
refsuite.parse(bpfile,fd=I.fd_emili,m2h=m2h,h2m=h2m,force_qualify=True,directed=True,\
force_score=0.0,qualify='bp')
bpfile.close()
def load_refsuite(filters='default',m2h=False,h2m=False,superdebug=False,force=False) :
from lib import interactors as I
from lib import interactors_extras as ie
global refsuite ;
if refsuite is not None and len(refsuite.nodes) > 0 and not force :
return ;
bgpath = REFERENCEPATH + 'biogrid_latest' ;
empath = REFERENCEPATH + '/complexes/emiliome.i' ;
bppath = REFERENCEPATH + '/bioplex.i' ;
mynf=None ;
myif=None ;
(myif,mynf)=ie.bg_regex_assembler() ;
if filters == 'default' :
pass ;
elif filters.lower() in 'exogenous' :
mynf=ie.exogenous_regex_assembler() ;
elif filters.lower() in 'ionly' :
mynf=None ;
elif filters is None or filters.lower() in 'none' :
mynf=None ;
myif=None ;
refsuite=I.dataSet(i_filter=myif,n_filter=mynf,superdebug=superdebug) ;
bgfile=open(bgpath) ;
refsuite.parse(bgfile,fd=I.fd_biogrid,m2h=m2h,h2m=h2m,force_qualify=True,\
force_score=0.0,qualify='bg') ;
bgfile.close() ;
emfile=open(empath) ;
refsuite.parse(emfile,fd=I.fd_emili,m2h=m2h,h2m=h2m,force_qualify=True,\
force_score=0.0,qualify='em') ;
emfile.close() ;
bpfile=open(bppath) ;
refsuite.parse(bpfile,fd=I.fd_emili,m2h=m2h,h2m=h2m,force_qualify=True,directed=True,\
force_score=0.0,qualify='bp') ;
bpfile.close() ;
def createNetwork(yamlfile):
readYAMLfile(yamlfile, config)
loadObjects(config)
theds = I.dataSet(n_filter=config['node_filter'])
readInDatasets(theds, config)
# filter experimental data by background dists
filterNodesByBackground(theds, config)
readPublicDatasets(theds, config)
secondaryFiltration(theds, config)
makeOutput(theds, config)
return theds, config
def readControls( c ):
cntrl = I.dataSet(n_filter = config['node_filter'], debug = DB)
for dsd in c['ds_dicts'] :
cfile = dsd['control_ori']
# assumed that this is the same in all datasets and
# it is a file of a list of ifilenames
break
c_dicts = list()
with open( cfile, 'rt' ) as fh:
for line in fh:
ifname = line.rstrip()
bait = line.split('_')[0]
c_dicts.append( { 'infilename': ifname, 'bait': bait, 'qualify': ifname })
c['c_dicts'] = c_dicts
readInDatasets( cntrl, c, 'c_dicts', 'cntrlkeys' )
return cntrl
def load_biogrid(filters='default',
m2h=False,
h2m=False,
force_qualify=True,
superdebug=False,
qualify=''):
global biogrid
bgpath = REFERENCEPATH + 'biogrid_latest'
from lib import interactors as I
from lib import interactors_extras as ie
mynf = None
myif = None
(myif, mynf) = ie.bg_regex_assembler()
if filters == 'default':
pass
elif filters.lower() in 'exogenous':
mynf = ie.exogenous_regex_assembler()
elif filters.lower() in 'ionly':
mynf = None
elif filters.lower() in 'physical':
physical_subf = ie.subfilter(r'genetic', 'systemType')
myif = ie.Filter(excludes={
physical_subf,
})
mynf = None
elif filters is None or filters.lower() in 'none':
mynf = None
myif = None
biogrid = I.dataSet(i_filter=myif, n_filter=mynf, superdebug=superdebug)
bgfile = open(bgpath)
biogrid.parse(bgfile,
fd=I.fd_biogrid,
m2h=m2h,
h2m=h2m,
force_qualify=force_qualify,
qualify=qualify)
bgfile.close()
def createNetwork( yamlfile ) :
readYAMLfile( yamlfile, config )
loadObjects( config )
theds = I.dataSet(n_filter = config['node_filter'])
readInDatasets( theds, config )
# filter experimental data by background dists
filterNodesByBackground( theds, config )
readPublicDatasets( theds, config )
secondaryFiltration( theds, config )
makeOutput( theds, config )
return theds,config
def createNetwork( yamlfile ) :
readYAMLfile( yamlfile, config )
loadObjects( config )
theds = I.dataSet(n_filter = config['node_filter'], debug = DB)
readInDatasets( theds, config, 'ds_dicts', 'baitkeys' )
if config['mt_method'] == 'fdr_bh':
# filter experimental data by background dists
filterNodesByBackground( theds, config )
elif config['mt_method'] == 'saintx':
scoreBySaintx( theds, config )
else:
config['joint_hits'] = set(theds.edges.keys())
config['node_pass1_all'] = set(theds.nodes.keys())
config['node_pass1_strong'] = set(theds.nodes.keys())
readPublicDatasets( theds, config )
if not config['rescueAll']:
secondaryFiltration( theds, config )
else :
config['nodes_pass2'] = config['node_pass1_all']
config['edges_pass1'] = set()
for e in list( theds.edges.values()) :
if {e.to.key,e.whence.key}.issubset( config['node_pass1_all'] ) and e.to.key != e.whence.key :
config['edges_pass1'].add(e)
config['edges_pass2'] = config['edges_pass1']
config['nnodes_rescued'] = 0
makeOutput( theds, config )
return theds,config
def readPublicDatasets( nwdata, c ):
if type(c['public_dicts']) is not list or len(c['public_dicts']) == 0:
return True
# OK, NOW we dump in the public datasets
for pd in c['public_dicts'] :
# (done) make emili follow biogrid field conventions
# TODO ditto bioplex
if pd['infilename'] in os.listdir('.') :
pdsf = open(pd['infilename'])
elif os.path.isfile(pd['infilename']) :
pdsf = open(pd['infilename'])
else :
pdsf = open(c['publicDatadir'] + pd['infilename'])
temporaryds = I.dataSet( i_filter = c['iact_filter'], debug = DB )
convert = pd.get('convert', None)
temporaryds.parse( pdsf, fd = I.fd_biogrid, convert = convert, qualify = pd.get('qualify',''),
directed = False, force_qualify = True, user = scruser )
sio = StringIO()
print( 'saving public dataset' + pd['infilename'])
# filter nodes of the dataset if bait is defined
if 'bait' in pd:
node_set = trim( temporaryds, pd.get('bait'), pd.get( 'radius', 1 ))
temporaryds.save( sio, nodes = node_set );
else :
edge_set = { e for e in temporaryds.edges.values() if e.weight >= pd.get('minweight',0) and e.totalscore >= pd.get('minscore',0) }
temporaryds.save( sio, edges = edge_set );
sio.seek(0)
print( 'import public data into network ' )
nwdata.load_from( sio, scruser )
sio.close()
pdsf.close()