def __init__(self,sample_sheet_file):
"""
Create a new SampleSheetBarcodes instance
Arguments:
sample_sheet_file (str): path of a SampleSheet.csv
file
"""
self._sample_sheet = SampleSheet(sample_sheet_file)
self._sample_lookup = {}
self._barcode_lookup = {}
self._lanes = []
sample_id = self._sample_sheet.sample_id_column
for line in self._sample_sheet.data:
if self._sample_sheet.has_lanes:
lane = line['Lane']
else:
lane = None
if lane not in self._lanes:
self._lanes.append(lane)
self._sample_lookup[lane] = {}
self._barcode_lookup[lane] = {}
sample = line[sample_id]
index_seq = normalise_barcode(samplesheet_index_sequence(line))
self._sample_lookup[lane][index_seq] = sample
self._barcode_lookup[lane][sample] = index_seq
def __init__(self, sample_sheet_file):
"""
Create a new SampleSheetBarcodes instance
Arguments:
sample_sheet_file (str): path of a SampleSheet.csv
file
"""
self._sample_sheet = SampleSheet(sample_sheet_file)
self._sample_lookup = {}
self._barcode_lookup = {}
self._lanes = []
sample_id = self._sample_sheet.sample_id_column
for line in self._sample_sheet.data:
if self._sample_sheet.has_lanes:
lane = line['Lane']
else:
lane = None
if lane not in self._lanes:
self._lanes.append(lane)
self._sample_lookup[lane] = {}
self._barcode_lookup[lane] = {}
sample = line[sample_id]
index_seq = samplesheet_index_sequence(line)
if index_seq is not None:
index_seq = index_seq.replace('-', '+')
else:
index_seq = ""
self._sample_lookup[lane][index_seq] = sample
self._barcode_lookup[lane][sample] = index_seq
def _check_sample_sheet_indexes(self,sample_sheet_file):
"""
Check that empty indexes are correctly specified in samplesheet
"""
# Split sample sheet into sub-sheets by lane
sample_sheet = SampleSheet(sample_sheet_file)
if sample_sheet.has_lanes:
lanes = list(set([line['Lane'] for line in sample_sheet]))
sample_sheet = [make_custom_sample_sheet(sample_sheet_file,
lanes=(i,))
for i in lanes]
else:
sample_sheet = [sample_sheet]
# Check for empty indexes in each lane
for s in sample_sheet:
for line in s:
if not samplesheet_index_sequence(line):
# Lane contains an empty index
# Only valid if this is the only line
if len(s.data) > 1:
if s.has_lanes:
raise Exception("Invalid sample sheet: "
"empty index must be the "
"only line for this lane")
else:
raise Exception("Invalid sample sheet: "
"empty index must be the "
"only line")
def get_bases_mask_icell8(bases_mask,sample_sheet=None):
"""
Reset the supplied bases mask string so that only the
bases containing the inline barcode and UMIs are kept,
and any remaining bases are ignored.
If a sample sheet is also supplied then an additional
update will be made to ensure that the bases mask
respects the barcode lengths given there.
Arguments:
bases_mask (str): initial bases mask string to update
sample_sheet (str): path to optional sample sheet
Returns:
String: updated bases mask string
"""
# Extract R1 mask
bases_mask = bases_mask.split(',')
r1_mask = bases_mask[0]
# Update to restrict to 21 bases
num_cycles = int(r1_mask[1:])
icell8_inline_length = (INLINE_BARCODE_LENGTH + UMI_LENGTH)
assert(num_cycles >= icell8_inline_length)
discard_length = (num_cycles - icell8_inline_length)
r1_mask = "y%d" % icell8_inline_length
r1_mask += ("n%d" % discard_length if discard_length > 0 else "")
bases_mask[0] = r1_mask
# Rebuild full bases mask
bases_mask = ','.join(bases_mask)
# Handle sample sheet
if sample_sheet is not None:
index_seq = samplesheet_index_sequence(
SampleSheet(sample_sheet).data[0])
if index_seq is None:
index_seq = ""
bases_mask = fix_bases_mask(bases_mask,index_seq)
return bases_mask
lanes=(lane, ))
else:
s = make_custom_sample_sheet(sample_sheet, fp.name)
if has_chromium_sc_indices(fp.name):
logging.warning("Lane %s has 10xGenomics Chromium "
"indices in sample sheet; not "
"matching against samplesheet for "
"this lane" % lane)
continue
# If mismatches not set then determine from
# the barcode lengths in the temporary
# samplesheet
if mismatches is None:
barcode_length = None
for line in s:
length = len(samplesheet_index_sequence(line))
if barcode_length is None:
barcode_length = length
elif length != barcode_length:
logging.error("Lane %s has a mixture of "
"barcode lengths" % lane)
barcode_length = min(barcode_length, length)
if barcode_length >= 6:
mismatches = 1
else:
mismatches = 0
# Check for collisions
while mismatches and check_barcode_collisions(
fp.name, mismatches):
mismatches = mismatches - 1
# Check mismatches
# Don't pass the sample sheet to the reporter
# for lanes with 10x indices
use_sample_sheet = None
logging.warning("Lane %s has 10xGenomics-style "
"indices in sample sheet; not "
"matching against samplesheet for "
"this lane" % lane)
else:
# If mismatches not set then determine from
# the barcode lengths in the temporary
# samplesheet
if mismatches is None:
barcode_length = None
for line in s:
index_sequence = \
samplesheet_index_sequence(line)
if index_sequence is None:
# Empty barcode sequence in
# samplesheet
length = 0
else:
length = len(index_sequence)
if barcode_length is None:
barcode_length = length
elif length != barcode_length:
logging.error("Lane %s has a mixture "
"of barcode lengths" % lane)
barcode_length = min(
barcode_length, length)
if barcode_length >= 6:
mismatches = 1