def test_generate_ind_template(self):
generate_ind_template(PATH_MOUSE_NOMENCLATURE, PATH_GENERIC_OUTPUT_TSV)
output = read_tsv(PATH_GENERIC_OUTPUT_TSV)
_label = 2
_description = 18
_aliases = 19
_rank = 20
self.assertTrue("PCL:" + get_individual_id("CS202002013_123")
in output) # child
test_node = output["PCL:" + get_individual_id("CS202002013_123")]
self.assertEqual("GABAergic", str(test_node[2]))
self.assertTrue(
str(test_node[_description]).startswith(
"GABAergic is: Neurons that use GABA as a neurotransmitter"))
self.assertEqual("Neuronal: GABAergic|Inhibitory neurons",
test_node[_aliases])
self.assertEqual("Class", test_node[_rank])
self.assertTrue("PCL:" + get_individual_id("CS202002013_219")
in output) # child
test_node = output["PCL:" + get_individual_id("CS202002013_219")]
self.assertEqual("Non-neural", str(test_node[2]))
self.assertTrue(
str(test_node[_description]).startswith(
"Non-Neural is: Cells of mesoderm"))
self.assertEqual("", test_node[_aliases])
self.assertEqual("Class", test_node[_rank])
self.assertEqual(
"Cell Type|Subclass",
output["PCL:" + get_individual_id("CS202002013_112")][_rank])
def generate_app_specific_template(taxonomy_file_path, output_filepath):
if str(taxonomy_file_path).endswith(".json"):
dend = dend_json_2_nodes_n_edges(taxonomy_file_path)
else:
dend = nomenclature_2_nodes_n_edges(taxonomy_file_path)
robot_template_seed = {
'ID': 'ID',
'TYPE': 'TYPE',
'cell_set_color': "A ALLENHELP:cell_set_color"
}
dl = [robot_template_seed]
for o in dend['nodes']:
if "cell_set_color" in o and o["cell_set_color"]:
d = dict()
d['ID'] = 'PCL:' + get_individual_id(o['cell_set_accession'])
d['TYPE'] = 'owl:NamedIndividual'
d['cell_set_color'] = str(o["cell_set_color"]).strip()
dl.append(d)
robot_template = pd.DataFrame.from_records(dl)
robot_template.to_csv(output_filepath, sep="\t", index=False)
def test_human_mtg_ids(self):
self.assertEqual(get_class_id("CS1908210001"), "0023001")
self.assertEqual(get_class_id("CS1908210148"), "0023148")
self.assertEqual(get_individual_id("CS1908210001"), "0023501")
self.assertEqual(get_individual_id("CS1908210148"), "0023648")
def test_marmoset_ids(self):
self.assertEqual(get_class_id("CS201912132_1"), "0019001")
self.assertEqual(get_class_id("CS201912132_121"), "0019121")
self.assertEqual(get_individual_id("CS201912132_1"), "0019501")
self.assertEqual(get_individual_id("CS201912132_121"), "0019621")
def test_human_ids(self):
self.assertEqual(get_class_id("CS201912131_1"), "0015001")
self.assertEqual(get_class_id("CS201912131_121"), "0015121")
self.assertEqual(get_individual_id("CS201912131_1"), "0015501")
self.assertEqual(get_individual_id("CS201912131_121"), "0015621")
def test_mouse_ids(self):
self.assertEqual(get_class_id("CS202002013_1"), "0011001")
self.assertEqual(get_class_id("CS202002013_121"), "0011121")
self.assertEqual(get_individual_id("CS202002013_1"), "0011501")
self.assertEqual(get_individual_id("CS202002013_121"), "0011621")
def generate_ind_template(taxonomy_file_path, output_filepath):
path_parts = taxonomy_file_path.split(os.path.sep)
taxon = path_parts[len(path_parts) - 1].split(".")[0]
if str(taxonomy_file_path).endswith(".json"):
dend = dend_json_2_nodes_n_edges(taxonomy_file_path)
else:
dend = nomenclature_2_nodes_n_edges(taxonomy_file_path)
taxon = path_parts[len(path_parts) - 1].split(".")[0].replace(
"nomenclature_table_", "")
dend_tree = generate_dendrogram_tree(dend)
taxonomy_config = read_taxonomy_config(taxon)
allen_descriptions = read_allen_descriptions(
ALLEN_DESCRIPTIONS_PATH, taxonomy_config['Species_abbv'][0])
subtrees = get_subtrees(dend_tree, taxonomy_config)
robot_template_seed = {
'ID': 'ID',
'Label': 'LABEL',
'PrefLabel': 'A skos:prefLabel',
'Entity Type': 'TI %',
'TYPE': 'TYPE',
'Property Assertions': "I 'subcluster of' SPLIT=|",
'Synonyms': 'A oboInOwl:hasExactSynonym SPLIT=|',
'Cluster_ID': "A 'cluster id'",
'Function': 'TI capable_of some %',
'cell_set_preferred_alias': "A n2o:cell_set_preferred_alias",
'original_label': "A n2o:original_label",
'cell_set_label': "A n2o:cell_set_label",
'cell_set_aligned_alias': "A n2o:cell_set_aligned_alias",
'cell_set_additional_aliases':
"A n2o:cell_set_additional_aliases SPLIT=|",
'cell_set_alias_assignee': "A n2o:cell_set_alias_assignee SPLIT=|",
'cell_set_alias_citation': "A n2o:cell_set_alias_citation SPLIT=|",
'Metadata': "A n2o:node_metadata",
'Exemplar_of': "TI 'exemplar data of' some %",
'Comment': "A rdfs:comment",
'Aliases': "A oboInOwl:hasRelatedSynonym SPLIT=|",
'Rank': "A 'cell_type_rank' SPLIT=|"
}
dl = [robot_template_seed]
synonym_properties = [
'cell_set_aligned_alias', 'cell_set_additional_aliases'
]
for o in dend['nodes']:
d = dict()
d['ID'] = 'PCL:' + get_individual_id(o['cell_set_accession'])
d['TYPE'] = 'owl:NamedIndividual'
d['Label'] = o['cell_set_label'] + ' - ' + o['cell_set_accession']
if 'cell_set_preferred_alias' in o and o['cell_set_preferred_alias']:
d['PrefLabel'] = o['cell_set_preferred_alias']
else:
d['PrefLabel'] = o['cell_set_accession']
d['Entity Type'] = 'PCL:0010001' # Cluster
d['Metadata'] = json.dumps(o)
d['Synonyms'] = '|'.join([
o[prop] for prop in synonym_properties
if prop in o.keys() and o[prop]
])
d['Property Assertions'] = '|'.join(
sorted([
'PCL:' + get_individual_id(e[1]) for e in dend['edges']
if e[0] == o['cell_set_accession']
]))
meta_properties = [
'cell_set_preferred_alias', 'original_label', 'cell_set_label',
'cell_set_aligned_alias', 'cell_set_additional_aliases',
'cell_set_alias_assignee', 'cell_set_alias_citation'
]
for prop in meta_properties:
if prop in o.keys():
d[prop] = '|'.join([
prop_val.strip() for prop_val in str(o[prop]).split("|")
if prop_val
])
else:
d[prop] = ''
d['Cluster_ID'] = o['cell_set_accession']
if o['cell_set_accession'] in set().union(
*subtrees) and o['cell_set_preferred_alias']:
d['Exemplar_of'] = PCL_BASE + get_class_id(o['cell_set_accession'])
if "cell_type_card" in o:
d['Rank'] = '|'.join([
cell_type.strip().replace("No", "None")
for cell_type in str(o["cell_type_card"]).split(",")
])
if o['cell_set_accession'] in allen_descriptions:
allen_data = allen_descriptions[o['cell_set_accession']]
d['Comment'] = allen_data["summary"][0]
if allen_data["aliases"][0]:
d['Aliases'] = '|'.join([
alias.strip()
for alias in str(allen_data["aliases"][0]).split("|")
])
# There should only be one!
dl.append(d)
robot_template = pd.DataFrame.from_records(dl)
robot_template.to_csv(output_filepath, sep="\t", index=False)
def generate_base_class_template(taxonomy_file_path, output_filepath):
taxon = extract_taxonomy_name_from_path(taxonomy_file_path)
taxonomy_config = read_taxonomy_config(taxon)
if taxonomy_config:
if str(taxonomy_file_path).endswith(".json"):
dend = dend_json_2_nodes_n_edges(taxonomy_file_path)
else:
dend = nomenclature_2_nodes_n_edges(taxonomy_file_path)
dend_tree = generate_dendrogram_tree(dend)
subtrees = get_subtrees(dend_tree, taxonomy_config)
if "Reference_gene_list" in taxonomy_config:
gene_db_path = ENSEMBLE_PATH.format(
str(taxonomy_config["Reference_gene_list"][0]).strip().lower())
gene_names = read_gene_data(gene_db_path)
minimal_markers = read_markers(
MARKER_PATH.format(taxon.replace("CCN", "").replace("CS", "")),
gene_names)
allen_markers = read_markers(
ALLEN_MARKER_PATH.format(
taxon.replace("CCN", "").replace("CS", "")), gene_names)
else:
minimal_markers = {}
allen_markers = {}
class_seed = [
'defined_class', 'prefLabel', 'Alias_citations',
'Synonyms_from_taxonomy', 'Gross_cell_type', 'Taxon',
'Brain_region', 'Minimal_markers', 'Allen_markers', 'Individual',
'Brain_region_abbv', 'Species_abbv', 'Cluster_ID', 'part_of',
'has_soma_location', 'aligned_alias', 'marker_gene_set'
]
class_template = []
for o in dend['nodes']:
if o['cell_set_accession'] in set.union(
*subtrees) and (o['cell_set_preferred_alias']
or o['cell_set_additional_aliases']):
d = dict()
d['defined_class'] = PCL_BASE + get_class_id(
o['cell_set_accession'])
if o['cell_set_preferred_alias']:
d['prefLabel'] = o['cell_set_preferred_alias']
elif o['cell_set_additional_aliases']:
d['prefLabel'] = str(
o['cell_set_additional_aliases']).split(
EXPRESSION_SEPARATOR)[0]
d['Synonyms_from_taxonomy'] = get_synonyms_from_taxonomy(o)
d['Gross_cell_type'] = get_gross_cell_type(
o['cell_set_accession'], subtrees, taxonomy_config)
d['Taxon'] = taxonomy_config['Species'][0]
d['Brain_region'] = taxonomy_config['Brain_region'][0]
d['Cluster_ID'] = o['cell_set_accession']
if 'cell_set_alias_citation' in o and o[
'cell_set_alias_citation']:
alias_citations = [
citation.strip() for citation in str(
o["cell_set_alias_citation"]).split("|")
if citation and citation.strip()
]
d["Alias_citations"] = "|".join(alias_citations)
if o['cell_set_accession'] in minimal_markers:
d['Minimal_markers'] = minimal_markers[
o['cell_set_accession']]
if o['cell_set_accession'] in allen_markers:
d['Allen_markers'] = allen_markers[o['cell_set_accession']]
else:
d['Allen_markers'] = ''
if 'Brain_region_abbv' in taxonomy_config:
d['Brain_region_abbv'] = taxonomy_config[
'Brain_region_abbv'][0]
if 'Species_abbv' in taxonomy_config:
d['Species_abbv'] = taxonomy_config['Species_abbv'][0]
d['Individual'] = PCL_BASE + get_individual_id(
o['cell_set_accession'])
for index, subtree in enumerate(subtrees):
if o['cell_set_accession'] in subtree:
location_rel = taxonomy_config['Root_nodes'][index][
'Location_relation']
if location_rel == "part_of":
d['part_of'] = taxonomy_config['Brain_region'][0]
d['has_soma_location'] = ''
elif location_rel == "has_soma_location":
d['part_of'] = ''
d['has_soma_location'] = taxonomy_config[
'Brain_region'][0]
if "cell_set_aligned_alias" in o and o[
"cell_set_aligned_alias"]:
d['aligned_alias'] = o["cell_set_aligned_alias"]
if o['cell_set_accession'] in minimal_markers:
d['marker_gene_set'] = PCL_PREFIX + get_marker_gene_set_id(
o['cell_set_accession'])
for k in class_seed:
if not (k in d.keys()):
d[k] = ''
class_template.append(d)
class_robot_template = pd.DataFrame.from_records(class_template)
class_robot_template.to_csv(output_filepath, sep="\t", index=False)
