def _run(obj):
"""Test reducing annotations as specified by the user"""
#pylint: disable=superfluous-parens
print('\nTESTING: {T} --------------------------------------------------'.
format(T=obj.name))
chk = _Chk(obj)
obj.chk_qualifiers()
if obj.name != 'id2gos':
assert chk.num_nots != 0, '{NAME} HAS 0 NOT Qualifiers'.format(
NAME=obj.name)
assc = obj.associations
num_assc = len(assc)
tic = timeit.default_timer()
_ev = obj.evobj
assc_dflt = obj.reduce_annotations(assc, AnnoOptions(_ev))
tic = obj.hms('Default', tic)
assc_nd0_not0 = obj.reduce_annotations(
assc, AnnoOptions(_ev, keep_ND=False, keep_NOT=False))
tic = obj.hms('{N:6,} assc_nd0_not0'.format(N=len(assc_nd0_not0)), tic)
assc_nd0_not1 = obj.reduce_annotations(
assc, AnnoOptions(_ev, keep_ND=False, keep_NOT=True))
tic = obj.hms('{N:6,} assc_nd0_not1'.format(N=len(assc_nd0_not1)), tic)
assc_nd1_not0 = obj.reduce_annotations(
assc, AnnoOptions(_ev, keep_ND=True, keep_NOT=False))
tic = obj.hms('{N:6,} assc_nd1_not0'.format(N=len(assc_nd1_not0)), tic)
assc_nd1_not1 = obj.reduce_annotations(
assc, AnnoOptions(_ev, keep_ND=True, keep_NOT=True))
tic = obj.hms('{N:6,} assc_nd1_not1'.format(N=len(assc_nd1_not1)), tic)
_red = obj.reduce_annotations
inc = set(_ev.code2nt.keys()).difference({'IEA'})
# pylint: disable=line-too-long
assert _red(assc, AnnoOptions(obj.evobj, ev_exclude={'IEA'})) == _red(
assc, AnnoOptions(obj.evobj, ev_include=inc))
inc = _ev.grp2codes['High_Throughput']
assert _red(assc, AnnoOptions(obj.evobj,
ev_include={'High_Throughput'})) == _red(
assc,
AnnoOptions(obj.evobj, ev_include=inc))
assert len(assc_dflt) == len(assc_nd0_not0)
assert len(assc_nd1_not1) == num_assc
# if obj.name in {'gaf', 'id2gos', 'gpad', 'id2go'}:
assert len(
assc_nd1_not0
) == num_assc - chk.num_nots, '{N}: ACT({A:,}) != EXP({E:,})'.format(
N=obj.name, A=len(assc_nd1_not0), E=num_assc - chk.num_nots)
assert len(
assc_nd0_not1
) == num_assc - chk.num_nds, '{N}: ACT({A:,}) != EXP({E:,})'.format(
N=obj.name, A=len(assc_nd0_not1), E=num_assc - chk.num_nds)
def _get_id2gos(self, associations, **kws):
"""Return given associations in a dict, id2gos"""
options = AnnoOptions(self.evobj, **kws)
# Default reduction is to remove. For all options, see goatools/anno/opts.py:
# * Evidence_Code == ND -> No biological data No biological Data available
# * Qualifiers contain NOT
assc = self.reduce_annotations(associations, options)
return self.get_dbid2goids(
assc) if options.b_geneid2gos else self.get_goid2dbids(assc)
def get_taxid2asscs(self, taxids=None, **kws):
"""Read Gene Association File (GAF). Return data."""
# WAS: get_annotations_taxid2dct
taxid2asscs = cx.defaultdict(lambda: cx.defaultdict(lambda: cx.defaultdict(set)))
options = AnnoOptions(self.evobj, **kws)
for taxid in self._get_taxids(taxids):
nts = self.taxid2asscs[taxid]
assc = self.reduce_annotations(nts, options)
taxid2asscs[taxid]['ID2GOs'] = self.get_dbid2goids(assc)
taxid2asscs[taxid]['GO2IDs'] = self.get_goid2dbids(assc)
return taxid2asscs
def _get_id2gos(self, associations, **kws):
"""Return given associations in a dict, id2gos"""
options = AnnoOptions(self.evobj, **kws)
# Default reduction is to remove. For all options, see goatools/anno/opts.py:
# * Evidence_Code == ND -> No biological data No biological Data available
# * Qualifiers contain NOT
assc = self.reduce_annotations(associations, options)
a2bs = self.get_dbid2goids(
assc) if options.b_geneid2gos else self.get_goid2dbids(assc)
# if not a2bs:
# raise RuntimeError('**ERROR: NO ASSOCATIONS FOUND: {FILE}'.format(FILE=self.filename))
return a2bs
def _get_id2gos(self, ntannos_usr, propagate_counts=False, relationships=None, prt=sys.stdout, **kws):
"""Return given ntannos_usr in a dict, id2gos"""
options = AnnoOptions(self.evobj, **kws)
# Default reduction is to remove. For all options, see goatools/anno/opts.py:
# * Evidence_Code == ND -> No biological data No biological Data available
# * Qualifiers contain NOT
ntannos_m = self.reduce_annotations(ntannos_usr, options)
dbid2goids = self.get_dbid2goids(ntannos_m, propagate_counts, relationships, prt)
if options.b_geneid2gos:
return dbid2goids
# if not a2bs:
# raise RuntimeError('**ERROR: NO ASSOCATIONS FOUND: {FILE}'.format(FILE=self.filename))
return self._get_goid2dbids(dbid2goids)
def _tst_ns2(obj, idx):
"""Test functions which use ns2 functions."""
# ALL annotations for a species
nts_all = obj.get_associations()
num_nts_all = len(nts_all)
num_nts_act = 0
# Separate ALL annotations into BP MF CC
ns2ntsanno = obj.get_ns2ntsanno()
assert set(ns2ntsanno.keys()).issubset({'BP', 'MF',
'CC'}), ns2ntsanno.keys()
# Reduce annotations to remove IEA
ns2anno_exp = {}
kws = {'ev_include': INC_GOOD}
for nspc, nts_orig in sorted(ns2ntsanno.items()):
opt = AnnoOptions(obj.evobj, **kws)
nts_redu = obj.reduce_annotations(nts_orig, opt)
num_nts_orig = len(nts_orig)
# Check that only current namespace is seen on namedtuples
assert set(nt.NS for nt in nts_orig if nt.NS == nspc) == {nspc}
num_nts_act += num_nts_orig
num_nts_redu = len(nts_redu)
# pylint: disable=line-too-long
print(
'{OPT} {IDX}) ns2ntanno {NS} {ALL:7,}=Loaded -> {N:7,} -> {R:7,} annos: {TYPE}'
.format(OPT=opt,
IDX=idx,
NS=nspc,
ALL=num_nts_all,
N=num_nts_orig,
R=num_nts_redu,
TYPE=obj.get_desc()))
ns2anno_exp[nspc] = obj.get_dbid2goids(nts_redu)
assert num_nts_all >= num_nts_orig
if obj.name == 'id2gos':
assert num_nts_orig == num_nts_redu
else:
assert num_nts_orig > num_nts_redu
assert num_nts_all >= num_nts_act
# Compare step-by-step transformation with top-level function, id2gos
ns2anno_act = obj.get_ns2assc(**kws)
for nspc, anno_exp in ns2anno_exp.items():
anno_act = ns2anno_act[nspc]
assert set(anno_exp.keys()) == set(anno_act.keys())
for geneid, gos_exp in anno_exp.items():
gos_act = anno_act[geneid]
assert gos_exp == gos_act
def read_ncbi_gene2go(fin_gene2go, taxids=None, **kws):
"""Read NCBI's gene2go. Return gene2go data for user-specified taxids."""
obj = Gene2GoReader(fin_gene2go, taxids)
# b_geneid2gos = not kws.get('go2geneids', False)
opt = AnnoOptions(**kws)
# By default, return id2gos. User can cause go2geneids to be returned by:
# >>> read_ncbi_gene2go(..., go2geneids=True
if 'taxid2asscs' not in kws:
if len(obj.taxid2asscs) == 1:
taxid = next(iter(obj.taxid2asscs))
return obj.get_annotations_dct(taxid, opt)
# Optional detailed associations split by taxid and having both ID2GOs & GO2IDs
# e.g., taxid2asscs = defaultdict(lambda: defaultdict(lambda: defaultdict(set))
t2asscs_ret = obj.get_annotations_taxid2dct(opt)
t2asscs_usr = kws.get(
'taxid2asscs',
defaultdict(lambda: defaultdict(lambda: defaultdict(set))))
if 'taxid2asscs' in kws:
obj.fill_taxid2asscs(t2asscs_usr, t2asscs_ret)
return obj.get_id2gos_all(t2asscs_ret)