def __init__(self):
self.test_name = args.test_name
self.sample_sets = GMT(args.gene_sets)
self.anno_list = GMT(args.annotation_list)
self.background = BACKGROUND([], args.background_list)
self.alpha = args.rate
self.output = args.output
self.cpu_count = args.cpu
self.precision = args.precision
self.console = str2bool(args.console)
self.significant = str2bool(args.significant)
self.table = str2bool(args.table)
def test_enrichment_score(self):
anno = GMT(os.path.join("files", "unittest_files",
"GO_shortened.gmt")).genesets['GO:0070507']
expr_list = MAT(os.path.join("files", "unittest_files",
"CLUSTERS.mat"))
cluster = 0
self.assertAlmostEqual(float(
enrichment_score(anno, cluster, expr_list, 1)),
0.629079429538,
delta=0.0001)
def test_page(self):
anno = GMT(os.path.join("files", "unittest_files", "GO_shortened.gmt"))
expr_list = MAT(os.path.join("files", "unittest_files",
"CLUSTERS.mat"))
cluster = 0
page_result = page(expr_list, cluster, anno, 1, cpu_count())
self.assertAlmostEqual(float(page_result[0][0].p_value),
6.35801240607e-17,
delta=0.0001)
self.assertAlmostEqual(float(page_result[0][1].p_value),
1.1116067475e-10,
delta=0.0001)
def test_wilcoxon(self):
anno = GMT(os.path.join("files", "unittest_files", "GO_shortened.gmt"))
expr_list = MAT(os.path.join("files", "unittest_files",
"CLUSTERS.mat"))
cluster = 0
wilcoxon_result = wilcoxon(expr_list, cluster, anno, 1, cpu_count())
self.assertAlmostEqual(float(wilcoxon_result[0][0].p_value),
1.1206994619e-10,
delta=0.0001)
self.assertAlmostEqual(float(wilcoxon_result[0][1].p_value),
0.0026701906744509285,
delta=0.0001)
def test_gsea(self):
anno = GMT(os.path.join("files", "unittest_files", "GO_shortened.gmt"))
expr_list = MAT(os.path.join("files", "unittest_files",
"CLUSTERS.mat"))
cluster = 0
permutations = 5
gsea_result = gsea(expr_list, cluster, anno, permutations, 1.0, 1,
cpu_count())
self.assertAlmostEqual(float(gsea_result[0][0].es),
0.629079429538,
delta=0.0001)
self.assertAlmostEqual(float(gsea_result[0][1].es),
0.458915389493,
delta=0.0001)
def test_generate_inputs(self):
anno = GMT(os.path.join("files", "unittest_files", "test_go.gmt"))
expr_list = MAT(os.path.join("files", "unittest_files",
"CLUSTERS.mat"))
cluster = 0
permutations = 5
test_set = list(range(0, 40))
test_set = set([str(i) for i in test_set])
inputs = generate_inputs(anno, cluster, expr_list, permutations)
self.assertEqual(inputs[0].anno_id, '0')
self.assertEqual(inputs[0].anno_list, test_set)
self.assertEqual(inputs[0].expr_cluster, 0)
self.assertEqual(inputs[0].expr_list, expr_list)
self.assertEqual(inputs[0].permutations, 5)
def __init__(self):
self.test_name = args.test_name
self.anno_list = GMT(args.annotation_list)
self.expr_list = MAT(args.expr_list)
#self.expr_list.DAVID_to_gene_symbol("C:\Users\Jimmy\Documents\dev\projects\enrichments\\files\\test_files\GPL_converted_list")
#self.expr_list.normalize(args.cluster_number)
self.expr_clusters = args.cluster_array
self.permutations = args.permutations
self.alpha = args.rate
self.output = args.output
self.weight = args.weight
self.cpu_count = args.cpu
self.precision = args.precision
self.console = str2bool(args.console)
self.significant = str2bool(args.significant)
self.table = str2bool(args.table)
nargs=1,
help='Ontology to use for propagation')
args = parser.parse_args()
MIN_POS, MAX_POS = 5, 500
if args.ontology == 'DO':
onto = DiseaseOntology.generate()
elif args.ontology == 'GO':
onto = GeneOntology.generate()
else:
onto = Ontology.generate()
if args.gmt:
# Load GMT genes onto Disease Ontology and propagate
gmt = GMT(filename=args.gmt)
onto.populate_annotations_from_gmt(gmt)
onto.propagate()
# Filter terms by number of gene annotations
terms = [
term.go_id for term in onto.get_termobject_list() if
len(term.annotations) >= MIN_POS and len(term.annotations) <= MAX_POS
]
if args.slim:
# Build ontology aware labels
lines = open(args.slim).readlines()
slim_terms = set([l.strip() for l in lines])
labels = OntoLabels(obo=onto, slim_terms=slim_terms)
else:
"--When filtering by publication, must provide GO namespace.\n")
sys.exit()
id_name = None
if args.idfile is not None:
id_name = IDMap(args.idfile)
gene_ontology = OBO(args.obo)
logger.info('Populating gene associations')
if args.ass:
gene_ontology.populate_annotations(args.ass,
gene_col=args.gcol,
term_col=args.term_col)
elif args.gmt:
gmt = GMT(args.gmt)
gene_ontology.populate_annotations_from_gmt(gmt)
else:
sys.stderr.write(
"--Provide gene annotations from an association file or a GMT file")
exit()
if args.pub_filter:
pub_counts = defaultdict(set)
for (term_id, term) in gene_ontology.go_terms.iteritems():
if term.namespace != args.nspace:
continue
for a in term.annotations:
pub_counts[a.ref].add((term, a))
for (ref, annots) in pub_counts.iteritems():
if len(annots) > 50:
args = parser.parse_args()
ubiq = None
if args.ubiq_genes:
ubiq = set()
with open(args.ubiq_genes) as f:
for l in f.readlines():
ubiq.add(l.strip())
logger.info('Total ubiquitous genes: %i', len(ubiq))
onto = None
if args.tissue_onto:
onto = OBO(args.tissue_onto)
tissue_genes = GMT(args.tissue_genes)
if onto:
onto.populate_annotations_from_gmt(tissue_genes)
onto.propagate()
tissue_genes = onto.as_gmt()
with open(args.pos) as f:
edge_lines = f.readlines()
tissue_std_edges = defaultdict(dict)
for line in edge_lines:
g1, g2, std = line.strip().split()[:3]
edge = frozenset([g1, g2])
type=str,
help='geneset id')
parser.add_argument('--prob',
'-p',
dest='prob_fit',
choices=['SIGMOID', 'ISO'],
default=None,
help='probability fit')
args = parser.parse_args()
standards = {}
Std = namedtuple('Std', ['pos', 'neg'])
if args.gmt:
gmt = GMT(filename=args.gmt)
if args.geneset_id:
pos_genes = gmt.get_genes(args.geneset_id)
neg_genes = gmt.genes - pos_genes
standards[args.geneset_id] = Std(pos=pos_genes, neg=neg_genes)
else:
for (gsid, genes) in gmt.genesets.iteritems():
pos_genes = gmt.get_genes(gsid)
neg_genes = gmt.genes - pos_genes
if len(pos_genes) >= 10 and len(pos_genes) <= 1000:
standards[gsid] = Std(pos=pos_genes, neg=neg_genes)
elif args.dir:
for f in os.listdir(args.dir):
pos_genes, neg_genes = set(), set()
with open(args.dir + '/' + f) as labelf:
lines = labelf.readlines()