def test_extract_protein_interactions_kgml(self, kgml_file,
expected_no_rel):
# Arrange
sut = KeggProteinInteractionsExtractor()
with open(
os.path.join(os.path.dirname(os.path.realpath(__file__)),
kgml_file), 'r') as myfile:
kgml_string = myfile.read()
# Mock Kegg ops
mock_kegg = KEGG()
sut.kegg = mock_kegg
# No matter what the input is, return the ko numbers that map to hsa numbers
mock_kegg.link = MagicMock(return_value="ko:K00922 hsa:5293\n" +
"ko:K00922 hsa:5291\n" +
"ko:K02649 hsa:5295")
# No matter what the input is, return the hsa numbers that map to uniprot numbers
mock_kegg.conv = MagicMock(return_value={"hsa:5293": "up:B0LPE5"})
# Mock Uni Prot
mock_uniprot = UniProt()
sut.uniprot = mock_uniprot
mock_uniprot.mapping = MagicMock(
return_value={"B0LPE5": ["gene1", "gene2"]})
# Act
actual = sut.extract_protein_interactions_kgml(kgml_string)
# Assert
self.assertEqual(expected_no_rel, len(actual))
def kegg_to_uniprot(fr='hsa', cache=False):
"""Downloads a mapping from a `KEGG` database to `UniProt`, including
both `TrEMBL` and `SwissProt`.
Parameters:
----------
fr : str, optional, default: 'hsa'
KEGG database identifier to convert. Defaults to 'hsa'.
cache : bool, optional, default: False
If True, results are cached by `bioservices`. This can save
time but you will eventually miss out on new database releases if
your cache is old.
Returns
-------
`dict`
Mapping from `KEGG` identifiers to a list of `UniProt` accessions.
"""
kegg = KEGG(cache=cache)
mapping = kegg.conv(fr, 'uniprot')
parsed_mapping = {}
for upid, org in mapping.items():
upid = upid.split(':')[1] # remove the 'up:' prefix
if org in parsed_mapping:
parsed_mapping[org] += [upid]
else:
parsed_mapping[org] = [upid]
return parsed_mapping
def tcell_read_metabolomics_data():
"""This function is quite convoluted as it downloads an excelfile from a publication and extracts a dataframe, idexed by chebi. The function also caches intermediate files"""
tcell_metabol_xls = cache.UrlFileCache(os.path.join(cache.get_cache_path(), metabolite_expression_name + ".xlsx"), metabolomics_data_url)
metabolomics_df = pd.read_excel(tcell_metabol_xls.get_file_name(), sheet_name = "normalized by sample mean", index_col=0, usecols="A,C:HN", skiprows = [0])
#metabolomics_df = pd.read_excel(tcell_metabol_xls.get_file_name(), sheet_name = "normalized by sample mean", index_col=0, usecols="A,C:HN", skiprows = [0])
for col in metabolomics_df.columns:
# Average all technical replicates (Named by trailing ".1")
if len(col.split('.'))>1 and col.split('.')[1] == "1":
remcol = col.split('.')[0]
metabolomics_df[remcol] = scipy.stats.gmean(metabolomics_df[[remcol,col]],axis=1)
metabolomics_df.drop(col, axis=1, inplace=True)
metabolomics_df.index.name = "KEGG_ID"
metabolomics_df = metabolomics_df.apply(np.exp2) # The excel data is in log2 space, return it to normal
k = KEGG(verbose=False)
map_kegg_chebi = k.conv("chebi", "compound")
metabolomics_df = metabolomics_df.groupby("KEGG_ID", group_keys=False).apply(lambda x: one_row_per_compound_convert(x, map_kegg_chebi)).reset_index(drop=True)
metabolomics_df.set_index("MetaboliteID", inplace=True)
return metabolomics_df
it.
'''
import os
import click
import json
import requests
import time
import xmltodict
import bioservices
from bioservices import KEGG, ChEBI
from zeep import Client
from tqdm import tqdm
k = KEGG(verbose=False)
map_kegg_chebi = k.conv("chebi", "compound")
c = ChEBI(verbose=False)
chebi_client = Client(
"https://www.ebi.ac.uk/webservices/chebi/2.0/webservice?wsdl")
chemspider_client = Client("https://www.chemspider.com/InChI.asmx?WSDL")
# For compounds that cant be found at all.
not_founds = []
# Need to create a global dictonary for these annotations, as I don't
# want to take the piss with the web services these wonderful people
# provide to us free of charge.
global CONVERTED_COMPOUNDS
CONVERTED_COMPOUNDS = {}
uniprot = UniProt(cache=True)
# ---- Set-up QuickGO bioservice
quickgo = QuickGO(cache=True)
# ---- Set-up KEGG bioservice
kegg, kegg_parser = KEGG(cache=True), KEGGParser()
kegg.organism = 'mmu'
print '[INFO] KEGG service configured'
kegg_pathways = {p: kegg.parse_kgml_pathway(p) for p in kegg.pathwayIds}
print '[INFO] KEGG pathways extracted: ', len(kegg_pathways)
# Convert KEGG pathways Gene Name to UniProt
k2u = kegg.conv('uniprot', 'mmu')
kegg_pathways_proteins = {p: {k2u[x].split(':')[1] for i in kegg_pathways[p]['entries'] if i['type'] == 'gene' for x in i['name'].split(' ') if x in k2u} for p in kegg_pathways}
kegg_uniprot_acc_map = {x for p in kegg_pathways_proteins for x in kegg_pathways_proteins[p]}
kegg_uniprot_acc_map = {p: uniprot.get_fasta(str(p)).split(' ')[0].split('|')[2] for p in kegg_uniprot_acc_map}
kegg_pathways_proteins = {p: {kegg_uniprot_acc_map[i] for i in kegg_pathways_proteins[p]} for p in kegg_pathways_proteins}
print '[INFO] KEGG pathways Ids converted to UniProt: ', len(kegg_pathways_proteins)
# ---- Set-up GO Terms gene list
go_terms_file = '%s/files/go_terms_uniprot.pickle' % wd
if os.path.isfile(go_terms_file):
with open(go_terms_file, 'rb') as handle:
go_terms = pickle.load(handle)
found = 0
found_list = []
missing = []
for line in ecorToSita:
ecor, sita = line.split(' ')
if ecor in ab_dict:
if len(ab_dict[ecor]) == 2:
ab_dict[ecor].append(sita)
found += 1
found_list.append(ecor)
else:
missing.append(ecor)
# load kegg module
from bioservices import KEGG
s = KEGG()
convDb = s.conv('sita', 'ncbi-proteinid')
convDb['ncbi-proteinid:YP_008815800']
# annotate kegg module
annotated = []
no_joy_for_sita = []
for gene in found_list:
sita = ab_dict[gene][-1]
sita_q = 'ncbi-proteinid:{g}'.format(g=sita[:-2])
if sita_q in convDb:
ab_dict[gene].append(convDb[sita_q])
annotated.append(gene)
else:
no_joy_for_sita.append(sita)
counter = 0
