def loadGTF(line, fp, fname, labels, regions, transcriptID, transcript_id_designator, defaultGroup):
"""
Like loadBED, but for a GTF file
This is largely a copy of what's in deeptoolsintervals
"""
file_label = dti.findRandomLabel(labels, os.path.basename(fname))
# handle the first line
cols = line.split("\t")
if cols[2].lower() == transcriptID.lower():
label, name = loadGTFtranscript(cols, file_label, defaultGroup, transcript_id_designator)
if label is not None:
if label not in labels:
labels[label] = len(labels)
regions.append(dict())
labelIdx = labels[label]
regions[labelIdx][name] = len(regions[labelIdx])
for line in fp:
if not isinstance(line, str):
line = line.decode('ascii')
if not line.startswith('#'):
cols = line.strip().split('\t')
if len(cols) == 0:
continue
if cols[2].lower() == transcriptID:
label, name = loadGTFtranscript(cols, file_label, defaultGroup, transcript_id_designator)
if label is None:
continue
if label not in labels:
labels[label] = len(labels)
regions.append(dict())
labelIdx = labels[label]
regions[labelIdx][name] = len(regions[labelIdx])
def loadBED(line, fp, fname, labelColumn, labels, regions, defaultGroup):
"""
Given a first line, possibly a label column and a list of labels and regions, add the labels and regions in the file to them
"""
# This is largely parseBED from deeptoolsintervals
labelIdx = None
localRegions = {}
cols = line.strip().split("\t")
if labelColumn is not None:
label = cols.pop(labelColumn)
if label not in labels:
labels[label] = len(labels)
labelIdx = labels[label]
if labelIdx >= len(regions):
regions.append(localRegions)
else:
localRegions = regions[labelIdx]
if len(cols) >= 6:
name = cols[3]
else:
name = "{0}:{1}-{2}".format(cols[0], cols[1], cols[2])
localRegions[name] = len(localRegions)
for line in fp:
if line.startswith("#") and labelColumn is None:
if len(localRegions) > 0:
label = line[1:].strip()
if len(label):
labels[dti.findRandomLabel(labels, label)] = len(labels)
else:
labels[dti.findRandomLabel(
labels, os.path.basename(fname))] = len(labels)
regions.append(localRegions)
localRegions = dict()
continue
elif line.startswith("#") and labelColumn is not None:
continue
cols = line.strip().split("\t")
if len(cols) < 3:
continue
if labelColumn is not None:
label = cols.pop(labelColumn)
if label not in labels:
labels[label] = len(labels)
labelIdx = labels[label]
if labelIdx >= len(regions):
regions.append({})
localRegions = regions[labelIdx]
if len(cols) >= 6:
name = cols[3]
else:
name = "{0}:{1}-{2}".format(cols[0], cols[1], cols[2])
name = dti.findRandomLabel(localRegions, name)
localRegions[name] = len(localRegions)
# Handle the last group if there is no label
if labelIdx is None and len(localRegions) > 0:
if defaultGroup is not None:
labels[dti.findRandomLabel(labels, defaultGroup)] = len(labels)
else:
labels[dti.findRandomLabel(labels,
os.path.basename(fname))] = len(labels)
regions.append(localRegions)
def loadBED(line, fp, fname, labelColumn, labels, regions, defaultGroup):
"""
Given a first line, possibly a label column and a list of labels and regions, add the labels and regions in the file to them
"""
# This is largely parseBED from deeptoolsintervals
labelIdx = None
localRegions = {}
cols = line.strip().split("\t")
if labelColumn is not None:
label = cols.pop(labelColumn)
if label not in labels:
labels[label] = len(labels)
labelIdx = labels[label]
if labelIdx >= len(regions):
regions.append(localRegions)
else:
localRegions = regions[labelIdx]
if len(cols) >= 6:
name = cols[3]
else:
name = "{0}:{1}-{2}".format(cols[0], cols[1], cols[2])
localRegions[name] = len(localRegions)
for line in fp:
if line.startswith("#") and labelColumn is None:
if len(localRegions) > 0:
label = line[1:].strip()
if len(label):
labels[dti.findRandomLabel(labels, label)] = len(labels)
else:
labels[dti.findRandomLabel(labels, os.path.basename(fname))] = len(labels)
regions.append(localRegions)
localRegions = dict()
continue
elif line.startswith("#") and labelColumn is not None:
continue
cols = line.strip().split("\t")
if len(cols) < 3:
continue
if labelColumn is not None:
label = cols.pop(labelColumn)
if label not in labels:
labels[label] = len(labels)
labelIdx = labels[label]
if labelIdx >= len(regions):
regions.append({})
localRegions = regions[labelIdx]
if len(cols) >= 6:
name = cols[3]
else:
name = "{0}:{1}-{2}".format(cols[0], cols[1], cols[2])
name = dti.findRandomLabel(localRegions, name)
localRegions[name] = len(localRegions)
# Handle the last group if there is no label
if labelIdx is None and len(localRegions) > 0:
if defaultGroup is not None:
labels[dti.findRandomLabel(labels, defaultGroup)] = len(labels)
else:
labels[dti.findRandomLabel(labels, os.path.basename(fname))] = len(labels)
regions.append(localRegions)
