def load_ensembl2interpro(self):
#Interpro
datafile = os.path.join(self.data_folder,
'gene_ensembl__prot_interpro__dm.txt')
for datadict in tab2dict_iter(datafile, (1, 4, 5, 6), 0):
datadict = dict_nodup(datadict)
# optimize with on call/convert
datadict = value_convert(
datadict, lambda x: {
'id': x[0],
'short_desc': x[1],
'desc': x[2]
})
# __aslistofdict__ : merge to 'interpro' as list of dict, not merging keys as list
# (these are merging instructions for later called merge_struct)
# 'interpro' : {'a': 1, 'b': 2} and 'interpro' : {'a': 3, 'b': 4} should result in:
# => 'interpro' : [{'a': 1, 'b': 2},{'a': 3, 'b': 4}]
# or not:
# => 'interpro' : {'a': [1,3], 'b': [2,4]}
datadict = value_convert(datadict,
lambda x: {
'interpro': x,
'__aslistofdict__': 'interpro'
},
traverse_list=False)
for doc in map_id(datadict, self.ensembl2entrez):
yield doc
def _load_ensembl_2taxid(self):
"""ensembl2taxid"""
datafile = os.path.join(self.data_folder,
'gene_ensembl__translation__main.txt')
ensembl2taxid = dict_nodup(
tab2dict(datafile, (0, 1), 1, includefn=_not_LRG))
# need to convert taxid to integer here
ensembl2taxid = value_convert(ensembl2taxid, lambda x: int(x))
return ensembl2taxid
def load_ensembl2pos(self):
datafile = os.path.join(
self.data_folder, 'gene_ensembl__gene__main.txt')
# Twice 1 because first is the dict key, the second because we need gene id within genomic_pos
ensembl2pos = dict_nodup(
tab2dict(datafile, (1, 1, 3, 4, 5, 6), 0, includefn=_not_LRG))
ensembl2pos = value_convert(ensembl2pos, lambda x: {
'ensemblgene': x[0], 'chr': x[3], 'start': int(x[1]), 'end': int(x[2]), 'strand': int(x[4])})
ensembl2pos = value_convert(ensembl2pos, lambda x: {
'genomic_pos': x}, traverse_list=False)
for datadict in tab2dict_iter(datafile, (1, 1, 3, 4, 5, 6), 0, includefn=_not_LRG):
datadict = dict_nodup(datadict)
datadict = value_convert(datadict, lambda x: {'ensemblgene': x[0], 'chr': x[3], 'start': int(
x[1]), 'end': int(x[2]), 'strand': int(x[4])})
datadict = value_convert(datadict, lambda x: {
'genomic_pos': x, '__aslistofdict__': 'genomic_pos'}, traverse_list=False)
for doc in map_id(datadict, self.ensembl2entrez):
yield doc
def load_ensembl2pfam(self):
# Prosite
datafile = os.path.join(
self.data_folder, 'gene_ensembl__prot_pfam__dm.txt')
for datadict in tab2dict_iter(datafile, (1, 4), 0):
datadict = dict_nodup(datadict)
datadict = value_convert(datadict, lambda x: {
'pfam': x}, traverse_list=False)
for doc in map_id(datadict, self.ensembl2entrez):
yield doc
def convert2entrez(self, ensembl2x):
'''convert a dict with ensembl gene ids as the keys to matching entrezgene ids as the keys.'''
if not self.ensembl2entrez_li:
self._load_ensembl2entrez_li()
if not self.ensembl_main:
self.ensembl_main = self.load_ensembl_main()
ensembl2entrez = list2dict(self.ensembl2entrez_li, 0)
entrez2ensembl = list2dict(self.ensembl2entrez_li, 1)
# Now make a dictionary indexed by entrez gene id
print('# of ensembl IDs in total: %d' %
len(set(ensembl2x) | set(ensembl2entrez)))
print('# of ensembl IDs match entrez Gene IDs: %d' %
len(set(ensembl2x) & set(ensembl2entrez)))
print('# of ensembl IDs DO NOT match entrez Gene IDs: %d' %
len(set(ensembl2x) - set(ensembl2entrez)))
# all genes with matched entrez
def _fn(eid, taxid=None):
# need to make a copy of the value here.
d = copy.copy(ensembl2x.get(eid, {}))
# otherwise, it will cause issue when multiple entrezgene ids
return d
# match the same ensembl gene, for example,
# ENSMUSG00000027104 --> (11909, 100047997)
data = value_convert(entrez2ensembl, _fn)
# add those has no matched entrez geneid, using ensembl id as the key
for eid in (set(ensembl2x) - set(ensembl2entrez)):
_g = ensembl2x[eid]
#_g.update(self.ensembl_main.get(eid, {}))
data[eid] = _g
for id in data:
if isinstance(data[id], dict):
_doc = dict_nodup(data[id], sort=True)
else:
# if one entrez gene matches multiple ensembl genes
_doc = dict_attrmerge(data[id], removedup=True, sort=True)
data[id] = _doc
return data
