def extract_Affy(fileName,
G,
refSet=None,
GO_columns=[30, 31, 32],
filetype="Affy",
delimiter=',',
quoting=csv.QUOTE_ALL):
fileName = checkForZip(fileName)
if (not os.path.exists(fileName)):
raise IOError(fileName + " does not exist and is required ")
#sniff and seek dialect
csvfile = readFile(fileName)
hasRef = None
if refSet:
hasRef = dict(izip(refSet, refSet))
GenetoGO, GOtoGene = dict(), dict()
for aspect in G.aspect:
GenetoGO[aspect], GOtoGene[aspect] = dict(), dict()
#Skip comments
row = csvfile.readline()
while row[0] == '#':
row = csvfile.readline()
#Read Header
header = row
#rd=list(csv.reader(f))
csv.register_dialect('format', delimiter=delimiter, quoting=quoting)
rd = csv.reader(csvfile, dialect='format')
for row in rd:
#Read gene product id
g = row[0]
if hasRef and not hasRef.has_key(g):
logger.handleWarning(
"gene product %s is not in the reference set, skip it " % g)
continue
for aspect, i in zip(
['biological_process', 'cellular_component', 'molecular_function'],
GO_columns):
for item in row[i].split('///'):
if not item == "---" and not len(item.strip()) == 0:
go = "GO:%07d" % int(item.split('//')[0].replace('/', ''))
go, aspect = G.get_GOAlternative(go, nameSpace=True)
if not aspect:
logger.handleWarning(
"term %s is not in GO graph, skip it " % go)
continue
GenetoGO[aspect].setdefault(g, set([])).add(go)
GOtoGene[aspect].setdefault(go, set([])).add(g)
return GenetoGO, GOtoGene
def extract_SCOP(fileName, G, refSet=None):
fileName = checkForZip(fileName)
if (not os.path.exists(fileName)):
raise IOError(fileName + " does not exist and is required ")
hasRef = None
if refSet:
hasRef = dict(izip(refSet, refSet))
GenetoGO, GOtoGene = dict(), dict()
for aspect in G.aspect:
GenetoGO[aspect], GOtoGene[aspect] = dict(), dict()
rd = csv.reader(readFile(fileName), delimiter=";")
header = rd.next()
for row in rd:
#Read gene product id
g = row[0]
g = row[header.index('domScop')]
go = row[header.index('termGo')]
if hasRef and not hasRef.has_key(g):
logger.handleWarning(
"gene product %s is not in the reference set, skip it " % g)
continue
if go.find('GO:') == 0:
#Get the alternative term if any and its GO aspect
go, aspect = G.get_GOAlternative(go, nameSpace=True)
if not aspect:
logger.handleWarning("term %s is not in GO graph, skip it " %
term)
continue
GenetoGO[aspect].setdefault(g, set([])).add(go)
GOtoGene[aspect].setdefault(go, set([])).add(g)
return GenetoGO, GOtoGene
def readGOoboXML(fileName, force=False, prefix="GO"):
import cPickle as pickle
picName = "%s.pic" % fileName
if (not os.path.exists(picName)):
force = True
if not force:
try:
logger.info("Reading serialized OBO file : %s" % picName)
with open(picName, "rb") as f:
G = pickle.load(f)
f.close()
except IOError as (inst):
print str(type(inst)) + " for " + picName
force = True
except EOFError as (inst): ##its an empty file?
print str(type(inst)) + " for " + picName
force = True
try:
if force:
fileName = checkForZip(fileName)
if (not os.path.exists(fileName)):
raise IOError(fileName + " does not exist and is required ")
logger.info("Reading OBO file : %s" % fileName)
G = get_GOGraph(readFile(fileName, mode="r"), prefix=prefix)
G.fileName = fileName
with open(picName, "wb") as f:
logger.info("Saving serialized OBO file")
pickle.dump(G, f, -1)
f.close()
except Exception, e:
logger.handleFatal("Unable to read file %s: %s" % (fileName, str(e)))
def readGAF_2(fileName):
GAF_col = [
"DB", "DB Object ID", "DB Object Symbol", "Qualifier", "GO ID",
"DB:Reference", "Evidence Code", "With (or) From", "Aspect",
"DB Object Name", "DB Object Synonym", "DB Object Type",
"Taxon(|taxon)", "Date", "Assigned By", "Annotation Extension",
"Gene Product Form ID"
]
#Read the entire file
data = [row for row in csv.reader(readFile(fileName), delimiter="\t")]
#Read the header
seek = 0
GAF_OK = False
while data[seek][0][0] == "!":
if re.search("!.*gaf-version.*:.*2", data[seek][0]):
GAF_OK = True
seek = seek + 1
if not GAF_OK:
raise Exception("Sorry, GAF format version 2.0 expected.")
return iter(data[seek:]), GAF_col
def add(self, fileName, refType="Fasta"):
if self.fileName == '':
self.fileName = fileName
self.refType = refType
else:
if type(self.fileName) == list:
self.fileName.append(fileName)
self.refType.append(refType)
else:
self.fileName = [self.fileName, fileName]
self.refType = [self.refType, refType]
fileName = checkForZip(fileName)
if (not os.path.exists(fileName)):
logger.handleFatal(fileName + " does not exist and is required ")
logger.info("Organism :\t%s" % self.organism)
logger.info("%s file :\t%s " % (refType, fileName))
try:
#Use fasta file to define the reference set
if refType == "Fasta":
from Bio import SeqIO
allID = set([
rec.name.split(";")[0].split(":")[-1]
for rec in SeqIO.parse(readFile(fileName), "fasta")
])
self.update(allID)
#Use a simple text file to define the reference set, first column is chosen by default
elif refType == "Text":
allID = set([
r[0] for r in csv.reader(readFile(fileName), delimiter=";")
])
self.update(allID)
#Use a GO annotation file to define the reference set
elif refType == "GAF":
from AIGO.IO import readGAF_2
data, GAF_col = readGAF_2(fileName)
allID = set([
".".join([
row[GAF_col.index("Taxon(|taxon)")][6:],
row[GAF_col.index("DB Object Symbol")]
]) for row in data
])
self.update(allID)
#Use a Affymetrix annotation file to define the reference set
elif refType == "AFFY":
f = readFile(fileName)
row = f.readline()
while row[0] == '#':
row = f.readline()
header = row
rd = csv.reader(f)
allID = set()
for row in rd:
#Read gene product id if not control sequence
if ("Control sequence".upper() != row[4].upper()):
allID.add(row[0])
self.update(allID)
else:
print "Sorry, unknown file type !!"
self.extend([])
raise Exception
if len(self) == 0:
logger.handleWarning("No gene products loaded")
except Exception, e:
logger.handleFatal("Unable to read file %s: %s" %
(fileName, str(e)))