def getKMERsForName(libname = 'mouse',tissue_term = None, **kwargs):
'''
Calls "parse" on the fasta file referred to as libname.
Optionally, specify a tissue term that will serve as a regex filter on the
FASTA record descriptors. Specifying a term such as "brain" will do
a case insensitive search on records in the library to return only kmers
from records matching "term".
'''
def setKMERsForName(**kwargs):
lname = kwargs['libname']
return parse( **kwargs)
name = libname if tissue_term ==None \
else '{0}_tissue={1}'.format(libname,tissue_term)
if kwargs.has_key('restored'):
output = kwargs['restored']
mem.getOrSet(setKMERsForName,
**mem.rc(kwargs,
libname = libname,
tissue_term = tissue_term,
name = name,
update = output))
else:
return mem.getOrSet(setKMERsForName,
**mem.rc(kwargs,
libname = libname,
tissue_term = tissue_term,
name = name))
def getTrackChrGenes(**kwargs):
'''
Get all of the genes from a bed file
on a given chromosome.
kwargs
num: chromosome number
fname: bedfile path (uses global bedfile as the default)
returns
a list of attributes for every gene.
'''
def setTrackChrGenes(**kwargs):
fname = kwargs.get('fname', bedfile)
num = kwargs.get('num', 1)
t = track.load(fname);
chromosome_data = t.read('chr{0}'.format(num))
rows = [dict(zip(r.keys(),r.data)) for r in iter(chromosome_data)]
return rows
return mem.getOrSet(setTrackChrGenes,
**mem.rc( kwargs,
onfail = 'compute',
name = '{0}_{1}'.format(kwargs.get('fname',os.path.basename(bedfile)),
kwargs.get('num', 1))
))
def getTranslatedForName(libname, **kwargs):
'''Translate kMERs to a numerical array for downstream analysis.'''
def setTranslatedForName(**kwargs):
global k
global translation
libname = kwargs.get('libname')
o = getKMERsForName( **mem.sr(kwargs, libname = libname))
translated = zeros((len(o),k))
idxed_mers = dict([(i,k) for i,k in enumerate(o.keys())])
occurrences=array([ o[idxed_mers[i]] for i in range(len(translated))])
d = translation
for i in idxed_mers.keys():
translated[i] = [d.get(l,4) for l in idxed_mers[i]]
return idxed_mers,translated, occurrences
tissue_term = kwargs.get('tissue_term', None)
name = libname if tissue_term ==None \
else '{0}_tissue={1}'.format(libname,tissue_term)
return mem.getOrSet(setTranslatedForName,
**mem.rc(kwargs,
libname = libname,
name = name))
def getBandCollectionAliases(**kwargs):
def setBandCollectionAliases(name = None, **kwargs):
assert name != None
all_aliases = []
freebase = discovery.build('freebase', 'v1', developerKey=DEVELOPER_KEY)
names_key = name
names_list = band_collectionnames[names_key]
for n in names_list:
q = [{
"name~=":"{0}".format(n),
"type": "/music/musical_group",
"/common/topic/alias": [{
"value": None
}],
"/music/musical_group/member": [{
"member": {
"/common/topic/alias": [{
"value": None
}]
}
}],
}]
responses = json.loads(freebase.mqlread(query=json.dumps(q)).execute())
for band in responses['result']:
member_aliases = [ a['value'] for e in band["/music/musical_group/member"] for a in e['member']["/common/topic/alias"]]
band_aliases = [a['value'] for a in band["/common/topic/alias"] ]
all_aliases.extend(member_aliases)
all_aliases.extend(band_aliases)
return all_aliases
name = kwargs['name']
return mem.getOrSet(setBandCollectionAliases, **mem.rc(kwargs,
name = name))
