def test_basic(my_go_term, my_other_term):
ontology = GeneOntology([my_go_term, my_other_term])
assert isinstance(ontology, GeneOntology)
assert isinstance(repr(ontology), str)
assert isinstance(str(ontology), str)
assert isinstance(text(ontology), text)
assert isinstance(ontology.hash, text)
# test access methods
assert len(ontology) == 2
assert my_go_term.id in ontology
assert ontology[my_go_term.id] == my_go_term
del ontology[my_go_term.id]
assert my_go_term.id not in ontology
ontology[my_go_term.id] = my_go_term
assert my_go_term.id in ontology
# test additional access methods
assert ontology.get_term_by_id(my_go_term.id) == my_go_term
assert ontology.get_term_by_acc(my_go_term.acc) == my_go_term
# test comparisons
other = copy.deepcopy(ontology)
assert other == ontology
del other[my_other_term.id]
assert other != ontology
# test iteration
assert set(list(iter(ontology))) == set([my_go_term, my_other_term])
def test_list(my_go_annotation, my_go_term):
gene_ontology = GeneOntology([my_go_term])
l = my_go_annotation.to_list()
assert isinstance(l, list)
other = GOAnnotation.from_list(gene_ontology, l)
assert isinstance(other, GOAnnotation)
assert other == my_go_annotation
def main(args=None):
"""Run GO-PCA and store the result in a `pickle` file.
Parameters
----------
args: argparse.Namespace object, optional
The argument values. If not specified, the values will be obtained by
parsing the command line arguments using the `argparse` module.
Returns
-------
int
Exit code (0 if no error occurred).
Raises
------
SystemError
If the version of the Python interpreter is not >= 2.7.
"""
vinfo = sys.version_info
if not (vinfo >= (2, 7)):
raise SystemError('Python interpreter version >= 2.7 required, '
'found %d.%d instead.' % (vinfo.major, vinfo.minor))
if args is None:
# read arguments from the command line
parser = get_argument_parser()
# parse first with default options, in case "--help" is specified
# ("--help" causes the program to exit at this point)
args = parser.parse_args()
# now remove the defaults and parse again
# (removing the defaults is important so that we know which values
# were specified by the user)
no_defaults = dict([p, None] for p in GOPCA.get_param_defaults())
no_defaults2 = dict([p, None] for p in GOPCAParams.get_param_defaults())
no_defaults.update(no_defaults2)
parser.set_defaults(**no_defaults)
args = parser.parse_args()
# reporting options
log_file = args.log_file
quiet = args.quiet
verbose = args.verbose
# test if we can write to log_file?
# configure root logger
logger = util.get_logger(log_file=log_file, quiet=quiet)
# check if required parameters were specified
passed = True
if args.expression_file is None:
logger.error('No expression file specified!')
passed = False
if args.gene_set_file is None:
logger.error('No gene set file specified!')
passed = False
if args.output_file is None:
logger.error('No output file specified!')
passed = False
if not passed:
logger.error('Not all required parameters were specified.')
return 1
# generate configuration
if args.config_file is not None:
# read parameter values from config file
params = GOPCAParams.read_ini(args.config_file)
else:
# start with default configuration
params = GOPCAParams()
# overwrite parameters specified on the command line
for p in GOPCAParams.get_param_defaults():
v = getattr(args, p)
if v is not None:
logger.debug('Parameter "%s" specified on command line!', p)
params.set_param(p, v)
global_params = GOPCA.get_param_defaults()
for k in list(global_params.keys()):
v = getattr(args, k)
if v is not None:
logger.debug('Parameter "%s" specified on command line!', p)
global_params[k] = v
# read expression file
matrix = ExpMatrix.read_tsv(args.expression_file)
logger.info('Expression matrix size: ' +
'(p = %d genes) x (n = %d samples).', matrix.p, matrix.n)
if args.sel_var_genes > 0:
# filter genes by variance
matrix = matrix.filter_variance(args.sel_var_genes)
# read gene set file
gene_sets = GeneSetCollection.read_tsv(args.gene_set_file)
print(args.gene_set_file, gene_sets)
# read ontology file (if supplied)
gene_ontology = None
if args.gene_ontology_file is not None:
p_logger = logging.getLogger(genometools.__name__)
p_logger.setLevel(logging.ERROR)
gene_ontology = GeneOntology.read_obo(
args.gene_ontology_file,
part_of_cc_only=params.go_part_of_cc_only)
p_logger.setLevel(logging.NOTSET)
M = GOPCA.simple_setup(matrix, params, gene_sets, gene_ontology,
verbose=verbose, **global_params)
run = M.run()
if run is None:
logger.error('GO-PCA run failed!')
return 1
# write run to pickle file
logger.info('Storing GO-PCA run in file "%s"...', args.output_file)
run.write_pickle(args.output_file)
return 0
def my_config(my_params, my_gene_ontology_file, my_fly_gene_set_file):
gene_ontology = GeneOntology.read_obo(my_gene_ontology_file)
gene_sets = GeneSetCollection.read_tsv(my_fly_gene_set_file)
config = GOPCAConfig(my_params, gene_sets, gene_ontology)
return config
def my_config(my_params, my_gene_ontology_file, my_fly_gene_set_file):
gene_ontology = GeneOntology.read_obo(my_gene_ontology_file)
gene_sets = GeneSetCollection.read_tsv(my_fly_gene_set_file)
config = GOPCAConfig(my_params, gene_sets, gene_ontology)
return config
def main(args=None):
"""Run GO-PCA and store the result in a `pickle` file.
Parameters
----------
args: argparse.Namespace object, optional
The argument values. If not specified, the values will be obtained by
parsing the command line arguments using the `argparse` module.
Returns
-------
int
Exit code (0 if no error occurred).
Raises
------
SystemError
If the version of the Python interpreter is not >= 2.7.
"""
vinfo = sys.version_info
if not (vinfo >= (2, 7)):
raise SystemError('Python interpreter version >= 2.7 required, '
'found %d.%d instead.' % (vinfo.major, vinfo.minor))
if args is None:
# read arguments from the command line
parser = get_argument_parser()
# parse first with default options, in case "--help" is specified
# ("--help" causes the program to exit at this point)
args = parser.parse_args()
# now remove the defaults and parse again
# (removing the defaults is important so that we know which values
# were specified by the user)
no_defaults = dict([p, None] for p in GOPCA.get_param_defaults())
no_defaults2 = dict([p, None]
for p in GOPCAParams.get_param_defaults())
no_defaults.update(no_defaults2)
parser.set_defaults(**no_defaults)
args = parser.parse_args()
# reporting options
log_file = args.log_file
quiet = args.quiet
verbose = args.verbose
# test if we can write to log_file?
# configure root logger
logger = util.get_logger(log_file=log_file, quiet=quiet)
# check if required parameters were specified
passed = True
if args.expression_file is None:
logger.error('No expression file specified!')
passed = False
if args.gene_set_file is None:
logger.error('No gene set file specified!')
passed = False
if args.output_file is None:
logger.error('No output file specified!')
passed = False
if not passed:
logger.error('Not all required parameters were specified.')
return 1
# generate configuration
if args.config_file is not None:
# read parameter values from config file
params = GOPCAParams.read_ini(args.config_file)
else:
# start with default configuration
params = GOPCAParams()
# overwrite parameters specified on the command line
for p in GOPCAParams.get_param_defaults():
v = getattr(args, p)
if v is not None:
logger.debug('Parameter "%s" specified on command line!', p)
params.set_param(p, v)
global_params = GOPCA.get_param_defaults()
for k in list(global_params.keys()):
v = getattr(args, k)
if v is not None:
logger.debug('Parameter "%s" specified on command line!', p)
global_params[k] = v
# read expression file
matrix = ExpMatrix.read_tsv(args.expression_file)
logger.info(
'Expression matrix size: ' + '(p = %d genes) x (n = %d samples).',
matrix.p, matrix.n)
if args.sel_var_genes > 0:
# filter genes by variance
matrix = matrix.filter_variance(args.sel_var_genes)
# read gene set file
gene_sets = GeneSetCollection.read_tsv(args.gene_set_file)
print(args.gene_set_file, gene_sets)
# read ontology file (if supplied)
gene_ontology = None
if args.gene_ontology_file is not None:
p_logger = logging.getLogger(genometools.__name__)
p_logger.setLevel(logging.ERROR)
gene_ontology = GeneOntology.read_obo(
args.gene_ontology_file, part_of_cc_only=params.go_part_of_cc_only)
p_logger.setLevel(logging.NOTSET)
M = GOPCA.simple_setup(matrix,
params,
gene_sets,
gene_ontology,
verbose=verbose,
**global_params)
run = M.run()
if run is None:
logger.error('GO-PCA run failed!')
return 1
# write run to pickle file
logger.info('Storing GO-PCA run in file "%s"...', args.output_file)
run.write_pickle(args.output_file)
return 0
def my_gene_ontology(my_gene_ontology_file):
gene_ontology = GeneOntology.read_obo(my_gene_ontology_file)
return gene_ontology
def my_gene_ontology(my_gene_ontology_file):
gene_ontology = GeneOntology.read_obo(my_gene_ontology_file)
return gene_ontology
def main(args=None):
"""Extract GO annotations and store in tab-delimited text file.
Parameters
----------
args: argparse.Namespace object, optional
The argument values. If not specified, the values will be obtained by
parsing the command line arguments using the `argparse` module.
Returns
-------
int
Exit code (0 if no error occurred).
Raises
------
SystemError
If the version of the Python interpreter is not >= 2.7.
"""
vinfo = sys.version_info
if not (vinfo >= (2, 7)):
raise SystemError('Python interpreter version >= 2.7 required, '
'found %d.%d instead.' % (vinfo.major, vinfo.minor))
if args is None:
parser = get_argument_parser()
args = parser.parse_args()
gene_file = args.gene_file
gene_ontology_file = args.gene_ontology_file
goa_association_file = args.goa_association_file
output_file = args.output_file
evidence_codes = args.evidence_codes
min_genes = args.min_genes_per_term
max_genes = args.max_genes_per_term
part_of_cc_only = args.part_of_cc_only
# logging parameters
log_file = args.log_file
quiet = args.quiet
verbose = args.verbose
# configure root logger
logger = misc.get_logger(log_file=log_file, quiet=quiet, verbose=verbose)
logger.info('Selected evidence codes: %s', ', '.join(evidence_codes))
logger.info('Min. number of genes per gene set: %d', min_genes)
logger.info('Max. number of genes per gene set: %d', max_genes)
# checks
assert os.path.isfile(gene_file)
assert os.path.isfile(gene_ontology_file)
assert os.path.isfile(goa_association_file)
# configure root logger
log_stream = sys.stdout
if output_file == '-':
# if we print output to stdout, redirect log messages to stderr
log_stream = sys.stderr
logger = misc.get_logger(log_stream=log_stream,
log_file=log_file,
quiet=quiet,
verbose=verbose)
# extract protein-coding genes from Ensembl GTF file
exp_genome = ExpGenome.read_tsv(gene_file)
# parse Gene Ontology
gene_ontology = GeneOntology.read_obo(gene_ontology_file)
# parse UniProt-GOA gene association file
with gzip.open(goa_association_file, 'rt', encoding='ascii') as fh:
go_annotations = ontology.parse_gaf(fh,
gene_ontology,
ev_codes=evidence_codes,
genome=exp_genome)
# extract GO-based gene sets
gene_sets = ontology.get_goa_gene_sets(go_annotations)
logger.info('Generated %d GO-derived gene sets', len(gene_sets))
# filter gene sets based on size
if min_genes > 0:
old_size = len(gene_sets)
gene_sets = GeneSetCollection(gs for gs in gene_sets
if gs.size >= min_genes)
logger.info('Excluded %d gene sets with too few genes.',
old_size - len(gene_sets))
if max_genes > 0:
old_size = len(gene_sets)
gene_sets = GeneSetCollection(gs for gs in gene_sets
if gs.size <= max_genes)
logger.info('Excluded %d gene sets with too many genes.',
old_size - len(gene_sets))
# writing output file
gene_sets.write_tsv(output_file)
logger.info('Wrote %s GO-derived gene sets to output file "%s".',
len(gene_sets), output_file)
return 0