def from_gene_list(args):
"""Creates gene signature from a pre-existing gene list, not expression
data.
"""
ranked_genes = args.get('ranked_genes', [])
gene_list = gene_list_factory.from_uploaded_ranked_genes(ranked_genes)
try:
required_metadata = required_metadata_factory.from_http_request(args)
except Exception as e:
raise HttpRequestArgumentsException('Error in required metadata '
'fields', e)
try:
optional_metadata = optional_metadata_factory.from_http_request(args)
except Exception as e:
raise HttpRequestArgumentsException('Error in optional metadata '
'fields', e)
try:
tags = tag_factory.from_http_request(args)
except Exception as e:
raise HttpRequestArgumentsException('Error in metadata tags', e)
resource = database.get_or_create(Resource,
code='upload',
name='User uploaded gene list')
return GeneSignature(None,
[gene_list],
required_metadata,
optional_metadata,
tags,
resource)
def from_gene_list(args):
"""Creates gene signature from a pre-existing gene list, not expression
data.
"""
ranked_genes = args.get('ranked_genes', [])
gene_list = gene_list_factory.from_uploaded_ranked_genes(ranked_genes)
try:
required_metadata = required_metadata_factory.from_http_request(args)
except Exception as e:
raise HttpRequestArgumentsException(
'Error in required metadata '
'fields', e)
try:
optional_metadata = optional_metadata_factory.from_http_request(args)
except Exception as e:
raise HttpRequestArgumentsException(
'Error in optional metadata '
'fields', e)
try:
tags = tag_factory.from_http_request(args)
except Exception as e:
raise HttpRequestArgumentsException('Error in metadata tags', e)
resource = database.get_or_create(Resource,
code='upload',
name='User uploaded gene list')
return GeneSignature(None, [gene_list], required_metadata,
optional_metadata, tags, resource)
def from_uploaded_ranked_genes(ranked_genes):
"""Creates GeneList from a list of lists, where the first element in the
inner list is the gene symbol and the second is the value.
"""
gene_list = [
RankedGene(database.get_or_create(Gene, name=g[0]), g[1])
for g in ranked_genes
]
direction = 0
target_app_links = []
return GeneList(gene_list, direction, target_app_links)
def from_uploaded_ranked_genes(ranked_genes):
"""Creates GeneList from a list of lists, where the first element in the
inner list is the gene symbol and the second is the value.
"""
gene_list = [
RankedGene(database.get_or_create(Gene, name=g[0]), g[1])
for g in ranked_genes
]
direction = 0
target_app_links = []
return GeneList(gene_list, direction, target_app_links)
def _create_gene_signature(soft_file, args):
"""Creates a new extraction, as opposed to an extraction from the
database.
"""
required_metadata = required_metadata_factory.from_http_request(args)
optional_metadata = optional_metadata_factory.from_http_request(args)
tags = tag_factory.from_http_request(args)
gene_lists = gene_list_factory.from_soft_file(soft_file, required_metadata,
optional_metadata, tags)
return GeneSignature(soft_file, gene_lists, required_metadata,
optional_metadata, tags,
database.get_or_create(Resource, code='geo'))
def from_soft_file(gene_signature):
"""Performs hierarchical clustering on SOFT file.
"""
target_app_link = __get_clustergrammer_link(gene_signature)
# Only create the link from Clustergrammer once.
# TODO: Move into targetapp module. API should not know about this.
if not target_app_link:
link = __from_soft_file(gene_signature)
link = '{0}?preview=true'.format(link)
target_app = database.get_or_create(TargetApp, name='clustergrammer')
target_app_link = TargetAppLink(target_app, link)
gene_signature.gene_lists[2].target_app_links.append(target_app_link)
database.save_gene_signature(gene_signature)
return target_app_link.link
def from_soft_file(gene_signature):
"""Performs hierarchical clustering on SOFT file.
"""
target_app_link = __get_clustergrammer_link(gene_signature)
# Only create the link from Clustergrammer once.
# TODO: Move into targetapp module. API should not know about this.
if not target_app_link:
link = __from_soft_file(gene_signature)
link = '{0}?preview=true'.format(link)
target_app = database.get_or_create(TargetApp, name='clustergrammer')
target_app_link = TargetAppLink(target_app, link)
gene_signature.gene_lists[2].target_app_links.append(
target_app_link
)
database.save_gene_signature(gene_signature)
return target_app_link.link
def _create_gene_signature(soft_file, args):
"""Creates a new extraction, as opposed to an extraction from the
database.
"""
required_metadata = required_metadata_factory.from_http_request(args)
optional_metadata = optional_metadata_factory.from_http_request(args)
tags = tag_factory.from_http_request(args)
gene_lists = gene_list_factory.from_soft_file(
soft_file,
required_metadata,
optional_metadata, tags
)
return GeneSignature(
soft_file,
gene_lists,
required_metadata,
optional_metadata,
tags,
database.get_or_create(Resource, code='geo')
)
def from_geo(args):
"""Constructs a SoftFile
"""
accession = args['dataset']
if not file_manager.file_exists(accession):
file_manager.download(accession)
dataset = database.get_geo_dataset(accession)
if not dataset:
platform = args['platform']
if platform.index('GPL') == 0:
platform = platform[3:]
title = args.get('title')
summary = args.get('summary')
if not title or not summary:
title, summary = eutils.get_title_and_summary(accession)
print('Fetched title and summary from EUtils API: %s, %s' %
(title, summary))
dataset = GeoDataset(accession=accession,
platform=platform,
organism=args['organism'],
title=title,
summary=summary)
else:
print('Dataset %s already exists!' % (accession))
a_cols = get_param_as_list(args, 'A_cols')
b_cols = get_param_as_list(args, 'B_cols')
# Use get_or_create to track GSMs. We don't do this for custom files.
control = [
database.get_or_create(SoftFileSample, name=sample, is_control=True)
for sample in a_cols
]
experimental = [
database.get_or_create(SoftFileSample, name=sample, is_control=False)
for sample in b_cols
]
samples = control + experimental
genes, a_vals, b_vals, selections, stats = parser.parse(
accession, True, dataset.platform, samples)
if 'normalize' not in args or args['normalize'] == 'True':
normalize = True
else:
normalize = False
genes, a_vals, b_vals = cleaner.clean(genes, a_vals, b_vals, normalize)
text_file = file_manager.write(accession, dataset.platform, normalize,
genes, a_vals, b_vals, samples, selections,
stats)
return SoftFile(samples,
dataset,
text_file,
genes,
a_vals,
b_vals,
stats=stats,
normalize=normalize)
def from_geo(args):
"""Constructs a SoftFile
"""
accession = args['dataset']
if not file_manager.file_exists(accession):
file_manager.download(accession)
dataset = database.get_geo_dataset(accession)
if not dataset:
platform = args['platform']
if platform.index('GPL') == 0:
platform = platform[3:]
title = args.get('title')
summary = args.get('summary')
if not title or not summary:
title, summary = eutils.get_title_and_summary(accession)
print('Fetched title and summary from EUtils API: %s, %s' % (title, summary))
dataset = GeoDataset(
accession=accession,
platform=platform,
organism=args['organism'],
title=title,
summary=summary
)
else:
print 'Dataset %s already exists!' % accession
a_cols = get_param_as_list(args, 'A_cols')
b_cols = get_param_as_list(args, 'B_cols')
# Use get_or_create to track GSMs. We don't do this for custom files.
control = [database.get_or_create(SoftFileSample,
name=sample,
is_control=True) for sample in a_cols]
experimental = [database.get_or_create(SoftFileSample,
name=sample,
is_control=False) for sample in b_cols]
samples = control + experimental
genes, a_vals, b_vals, selections, stats = parser.parse(accession,
True,
dataset.platform,
samples)
if 'normalize' not in args or args['normalize'] == 'True':
normalize = True
else:
normalize = False
genes, a_vals, b_vals = cleaner.clean(genes, a_vals, b_vals, normalize)
text_file = file_manager.write(accession, dataset.platform, normalize,
genes, a_vals, b_vals, samples, selections,
stats)
return SoftFile(
samples, dataset, text_file,
genes, a_vals, b_vals,
stats=stats, normalize=normalize
)