def random_foreground_background(
): # used TaxIDs fixture previously, but now it is random on TaxID level as well
for _ in range(10):
taxid = random.choice(query.get_taxids()) # read_from_flat_files=True
background = query.get_proteins_of_taxid(taxid)
foreground = random.sample(background, 200)
return foreground, background, taxid
def test_ENSP_consistency_of_DB():
"""
- ENSPs of taxid_2_protein_table are the superset of ENSPs of protein_2_function_table
foreground with functional association also has to be in the precomputed background
TaxID_2_Protein_table_STRING: ENSPs expected to be the superset of Protein_2_Function_table_STRING
Protein_2_Function_table_STRING
Function_2_ENSP_table_STRING
"""
for taxid in query.get_taxids():
ensp_taxid_2_protein = set(query.get_proteins_of_taxid(taxid))
ensp_protein_2_function = {
ele[0]
for ele in query.get_results_of_statement(
"SELECT protein_2_function.an FROM protein_2_function WHERE protein_2_function.an ~ '^{}\.'"
.format(taxid))
}
# ensp_function_2_ensp = None
len_ensp_taxid_2_protein = len(ensp_taxid_2_protein)
len_ensp_protein_2_function = len(ensp_protein_2_function)
assert len_ensp_taxid_2_protein >= len_ensp_protein_2_function
assert len(ensp_taxid_2_protein.intersection(
ensp_protein_2_function)) == len_ensp_protein_2_function
assert len(ensp_taxid_2_protein.union(
ensp_protein_2_function)) == len_ensp_taxid_2_protein
def test_run_STRING_enrichment(pqo_STRING, STRING_examples, args_dict):
"""
checking that
:param pqo_STRING: PersistentQuery Object
:param STRING_examples: tuple (foreground ENSPs, taxid)
:param args_dict: dict (from conftest.py with default values)
:return
:
"""
enrichment_method = "compare_samples"
foreground, taxid = STRING_examples
background = query.get_proteins_of_taxid(taxid)
# background_n = pqo_STRING.get_proteome_count_from_taxid(taxid)
args_dict_temp = args_dict.copy()
args_dict_temp.update({"foreground":format_for_REST_API(foreground),
"background":format_for_REST_API(background),
"intensity":None,
"enrichment_method":enrichment_method})
# ui = userinput.REST_API_input(pqo_STRING, foreground_string=format_for_REST_API(foreground),
# background_string=format_for_REST_API(background), background_intensity=None, enrichment_method=enrichment_method)
ui = userinput.REST_API_input(pqo_STRING, args_dict_temp)
# results_all_function_types = run.run_STRING_enrichment(pqo=pqo_STRING, ui=ui, enrichment_method=enrichment_method,
# limit_2_entity_type=variables.limit_2_entity_types_ALL, output_format="json", FDR_cutoff=None)
args_dict_temp.update({"limit_2_entity_type":variables.limit_2_entity_types_ALL,
"output_format":"json",
"FDR_cutoff":None})
results_all_function_types = run.run_STRING_enrichment(pqo=pqo_STRING, ui=ui, args_dict=args_dict_temp)
assert results_all_function_types != {'message': 'Internal Server Error'}
etypes = variables.entity_types_with_data_in_functions_table
assert len(set(results_all_function_types.keys()).intersection(etypes)) == len(etypes)
for _, result in results_all_function_types.items():
# assert result is not empty
assert result
def random_abundance_correction_foreground_background():
for _ in range(10):
taxid = random.choice(query.get_taxids()) # read_from_flat_files=True
background = query.get_proteins_of_taxid(taxid)
foreground = random.sample(background, 200)
intensity = [
str(ele) for ele in np.random.normal(size=len(background))
]
return foreground, background, intensity, taxid
def create_functions_2_ENSP_table():
for taxid in fileinput.input():
taxid = taxid.strip()
ans_list = sorted(query.get_proteins_of_taxid(taxid))
etype_2_association_dict = query.PersistentQueryObject_STRING.get_association_dict_split_by_category(ans_list)
for etype in sorted(variables.entity_types_with_data_in_functions_table):
assoc_dict = etype_2_association_dict[etype]
association_2_count_dict, association_2_ANs_dict, ans_counter = count_terms_v3(set(ans_list), assoc_dict)
for association, ans in association_2_ANs_dict.items():
assert ans_counter >= association_2_count_dict[association]
print(str(taxid) + "\t" + str(etype) + "\t" + association + "\t" + str(association_2_count_dict[association]) + "\t" + str(ans_counter) + "\t" + "{" + str(sorted(ans))[1:-1].replace(" ", "").replace("'", '"') + "}")
def test_precomputed_associations_counts(pqo_STRING, TaxIDs):
taxid = TaxIDs
ENSPs_proteome = query.get_proteins_of_taxid(taxid)
# A
etype_2_association_2_count_dict_background, etype_2_association_2_ANs_dict_background, etype_2_background_n = query.get_association_2_count_ANs_background_split_by_entity(
taxid)
etype_2_association_dict = pqo_STRING.get_association_dict_split_by_category(
set(ENSPs_proteome))
for etype in variables.entity_types_with_data_in_functions_table:
# B
association_2_count_dict, association_2_ANs_dict, ans_counter = ratio.count_terms_v3(
set(ENSPs_proteome), etype_2_association_dict[etype])
assert association_2_count_dict == etype_2_association_2_count_dict_background[
etype]
assert association_2_ANs_dict == etype_2_association_2_ANs_dict_background[
etype]
# UniProt_IDs_human_list = sorted(query.get_proteins_of_taxid(9606, read_from_flat_files=True))
# ENSP_human_list = sorted(query.get_proteins_of_human())
#
# @pytest.fixture(scope="session")
# def UniProt_IDs_human():
# return UniProt_IDs_human_list
#
# @pytest.fixture(scope="session")
# def ENSPs_human():
# return ENSP_human_list
### preloaded objects --> put into conftest.py?
UniProt_IDs_human_list = sorted(
query.get_proteins_of_taxid(9606, read_from_flat_files=True))
ENSP_human_list = sorted(query.get_proteins_of_human())
###
def get_random_human_ENSP(num_ENSPs=20,
joined_for_web=False,
contiguous=False,
UniProt_ID=False,
UniProt_IDs_human_list=UniProt_IDs_human_list,
ENSP_human_list=ENSP_human_list):
if UniProt_ID:
IDs_2_sample = UniProt_IDs_human_list
else:
IDs_2_sample = ENSP_human_list
max_index = len(IDs_2_sample)