def get_part_sequence(self, fasta_file, header, start, stop, nterminus, strand, name):
"""
Pull part sequence from fasta file
# https://github.com/mdshw5/pyfaidx
# pip install pyfaidx
Parameters
----------
Args:
fasta_file (str): input fasta file
header (str): header for fasta sequence
start (str): start coordinate
stop (str): stop coordinate
nterminus (int): length of missing sequence
strand (str): strand
name (str): gene name
Returns:
sequence (str): portion on a sequence
"""
# remove the last 2 characters from header as this is appended by prodigal
header = header[:header.rfind("_")]
# logger.info("[PARTIAL] ARO: {} | contig: {} | filename: {}".format(name, header, fasta_file))
genes = Fasta(fasta_file, sequence_always_upper=False, read_long_names=False, one_based_attributes=True)
# logger.info(genes.records)
logger.info(json.dumps({"strand":strand, "start":start, "stop":stop, "nterminus":nterminus}, indent=2))
if strand == "-":
return str(genes.get_spliced_seq( header, [[stop, stop+nterminus]]))
elif strand == "+":
return str(genes.get_spliced_seq( header, [[start-nterminus, start]]))
def test_split_seq(self):
""" Fetch sequence by blocks """
fa = Fasta('data/chr17.hg19.part.fa')
gene = Fasta("data/gene.bed12.fasta")
expect = gene[list(gene.keys())[0]][:].seq
bed = "data/gene.bed12"
with open(bed) as fi:
record = fi.readline().strip().split("\t")
chrom = record[0]
start = int(record[1])
strand = record[5]
# parse bed12 format
starts = [int(x) for x in record[11].split(",")[:-1]]
sizes = [int(x) for x in record[10].split(",")[:-1]]
starts = [start + x for x in starts]
ends = [start + size for start, size in zip(starts, sizes)]
# bed half-open
if strand == "-":
starts = [start + 1 for start in starts]
else:
ends = [end - 1 for end in ends]
intervals = zip(starts, ends)
result = fa.get_spliced_seq(chrom, intervals, rc=True)
print(result.seq)
print("====")
print(expect)
assert result.seq == expect
def test_split_seq(self):
""" Fetch sequence by blocks """
fa = Fasta('data/chr17.hg19.part.fa')
gene = Fasta("data/gene.bed12.fasta")
expect = gene[list(gene.keys())[0]][:].seq
bed = "data/gene.bed12"
with open(bed) as fi:
record = fi.readline().strip().split("\t")
chrom = record[0]
start = int(record[1])
strand = record[5]
# parse bed12 format
starts = [int(x) for x in record[11].split(",")[:-1]]
sizes = [int(x) for x in record[10].split(",")[:-1]]
starts = [start + x for x in starts]
ends = [start + size for start,size in zip(starts, sizes)]
# bed half-open
if strand == "-":
starts = [start + 1 for start in starts]
else:
ends = [end - 1 for end in ends]
intervals = zip(starts, ends)
result = fa.get_spliced_seq(chrom, intervals, rc=True)
print(result.seq)
print("====")
print(expect)
assert result.seq == expect
def get_part_sequence(self, fasta_file, header, start, stop, nterminus,
strand, name):
"""
Pull part sequence from fasta file
# https://github.com/mdshw5/pyfaidx
# pip install pyfaidx
Parameters
----------
Args:
fasta_file (str): input fasta file
header (str): header for fasta sequence
start (str): start coordinate
stop (str): stop coordinate
nterminus (int): length of missing sequence
strand (str): strand
name (str): gene name
Returns:
sequence (str): portion on a sequence
"""
# remove the last 2 characters from header as this is appended by prodigal
header = header[:header.rfind("_")]
genes = False
# logger.info("[PARTIAL] ARO: {} | contig: {} | filename: {}".format(name, header, fasta_file))
try:
genes = Fasta(fasta_file,
sequence_always_upper=False,
read_long_names=False,
one_based_attributes=True)
except Exception as e:
logger.error(e)
# logger.info(genes.records)
if genes:
# logger.debug(json.dumps({"strand":strand, "start":start, "stop":stop, "nterminus":nterminus}, indent=2))
if strand == "-":
_start = stop + 1
_stop = stop + nterminus
# logger.debug("grep sequence from {}|-|{}-{}".format(header,_start, _stop,))
if nterminus == 0:
# logger.debug("grep sequence from {}|-|{}-{}".format(header,start, stop,))
return str(genes.get_spliced_seq(
header, [[start, stop]])), start, stop
else:
return str(genes.get_spliced_seq(
header, [[_start, _stop]])), _start, _stop
elif strand == "+":
_start = start - nterminus
_stop = start - 1
if _start <= 0:
_start = 1
if _stop <= 0:
_stop = 1
# logger.debug("grep sequence from {}|+|{}-{}".format(header,_start, _stop))
if nterminus == 0:
# logger.debug("grep sequence from {}|+|{}-{}".format(header,start, stop))
return str(genes.get_spliced_seq(
header, [[start, stop]])), start, stop
else:
return str(genes.get_spliced_seq(
header, [[_start, _stop]])), _start, _stop
# cds = cds.sort_values(['chr','startGene'])
for index,row in cds.iterrows():
print(index,row['id'])
start = time.time()
# Convert CDS list into numeric array
coordinates = array(row['coordinates'].split(',')).astype(int).tolist()
coordinates = [coordinates[i:i+2] for i in range(0, len(coordinates), 2)]
# Open ref and outgroup
ref = Fasta('/data/shared/dgn/ref/Chr' + row['chr'] +'.fasta',sequence_always_upper=True)
outgroup = Fasta(outgroupFastas + '/Chr' + row['chr'] +'_dsim.fasta',sequence_always_upper=True)
## Extract ref and outgroup seq
refSeq = ref.get_spliced_seq(list(ref.keys())[0], coordinates).seq
outgroupSeq = outgroup.get_spliced_seq(list(outgroup.keys())[0], coordinates).seq
# Check length divisible by 3
if((len(refSeq) % 3) == 0):
# Open multifasta
multiFasta = Fasta('/data/shared/dgn/alignments/'+ args.population + '_Chr' + row['chr'] +'.seq',sequence_always_upper=True)
# Extract samples from fastas
samples = list(multiFasta.keys())
# Create empty array with ndimesions equal to multi-Fasta lines and length
matrix = np.empty([len(samples) + 1, len(refSeq)],dtype='str')
positions=[]
for index, row in cds.iterrows():
print(index, row['id'])
start = time.time()
# Convert CDS list into numeric array
coordinates = array(row['coordinates'].split(',')).astype(int).tolist()
coordinates = [
coordinates[i:i + 2] for i in range(0, len(coordinates), 2)
]
# Open ref and outgroup
ref = Fasta('/data/shared/dgn/ref/Chr' + row['chr'] + '.fasta')
outgroup = Fasta(outgroupFastas + '/Chr' + row['chr'] + '_dsim.fasta')
## Extract ref and outgroup seq
refSeq = ref.get_spliced_seq(row['chr'], coordinates).seq.upper()
outgroupSeq = outgroup.get_spliced_seq(outputHeader,
coordinates).seq.upper()
if ('M' in refSeq):
continue
else:
if ((len(refSeq) / 3).is_integer()):
# Open population multifasta
popFasta = Fasta('/data/shared/dgn/alignments/' +
args.population + '_Chr' + row['chr'] +
'.seq')
#Extract samples
samples = list(popFasta.keys())
