def main():
args = parse_args()
sys.stderr.write("command line: %s\n" % " ".join(sys.argv))
sys.stderr.write("python version: %s\n" % sys.version)
sys.stderr.write("pysam version: %s\n" % pysam.__version__)
sys.stderr.write("pytables version: %s\n" % tables.__version__)
util.check_pysam_version()
util.check_pytables_version()
# disable warnings that come from pytables when chromosome
# names are like 1, 2, 3 (instead of chr1, chr2, chr3)
warnings.filterwarnings('ignore', category=tables.NaturalNameWarning)
snp_tab_h5 = tables.open_file(args.snp_tab, "r")
snp_index_h5 = tables.open_file(args.snp_index, "r")
if args.haplotype:
hap_h5 = tables.open_file(args.haplotype, "r")
else:
hap_h5 = None
ref_count_h5 = tables.open_file(args.ref_as_counts, "w")
alt_count_h5 = tables.open_file(args.alt_as_counts, "w")
other_count_h5 = tables.open_file(args.other_as_counts, "w")
read_count_h5 = tables.open_file(args.read_counts, "w")
output_h5 = [ref_count_h5, alt_count_h5, other_count_h5, read_count_h5]
chrom_dict = {}
# initialize every chromosome in output files
chrom_list = chromosome.get_all_chromosomes(args.chrom)
for chrom in chrom_list:
for out_file in output_h5:
create_carray(out_file, chrom, args.data_type)
chrom_dict[chrom.name] = chrom
count = 0
dtype = None
if args.data_type == "uint8":
max_count = MAX_UINT8_COUNT
dtype = np.uint8
elif args.data_type == "uint16":
max_count = MAX_UINT16_COUNT
dtype = np.uint16
else:
raise NotImplementedError("unsupported datatype %s" % args.data_type)
# create a txt file to also holds the counts
if args.txt_counts is not None:
if os.path.splitext(args.txt_counts)[1] == ".gz":
txt_counts = gzip.open(args.txt_counts, 'wt+')
else:
txt_counts = open(args.txt_counts, 'w+')
for chrom in chrom_list:
sys.stderr.write("%s\n" % chrom.name)
if args.test_chrom:
if chrom.name != args.test_chrom:
sys.stderr.write("skipping because not test chrom\n")
continue
warned_pos = {}
# fetch SNP info for this chromosome
if chrom.name not in snp_tab_h5.root:
# no SNPs for this chromosome
sys.stderr.write("skipping %s because chromosome with this name "
"not found in SNP table\n" % chrom.name)
continue
sys.stderr.write("fetching SNPs\n")
snp_tab = snp_tab_h5.get_node("/%s" % chrom.name)
snp_index_array = snp_index_h5.get_node("/%s" % chrom.name)[:]
if hap_h5:
hap_tab = hap_h5.get_node("/%s" % chrom.name)
ind_dict, ind_idx = snptable.SNPTable().get_h5_sample_indices(
hap_h5, chrom, [args.individual])
if len(ind_idx) == 1:
ind_idx = ind_idx[0]
sys.stderr.write("index for individual %s is %d\n" %
(args.individual, ind_idx))
else:
raise ValueError("got sample indices for %d individuals, "
"but expected to get index for one "
"individual (%s)" %
(len(ind_idx), args.individual))
hap_tab = None
ind_idx = None
else:
hap_tab = None
ind_idx = None
# initialize count arrays for this chromosome to 0
ref_carray = get_carray(ref_count_h5, chrom)
alt_carray = get_carray(alt_count_h5, chrom)
other_carray = get_carray(other_count_h5, chrom)
read_count_carray = get_carray(read_count_h5, chrom)
ref_array = np.zeros(chrom.length, dtype)
alt_array = np.zeros(chrom.length, dtype)
other_array = np.zeros(chrom.length, dtype)
read_count_array = np.zeros(chrom.length, dtype)
# loop over all BAM files, pulling out reads
# for this chromosome
for bam_filename in args.bam_filenames:
sys.stderr.write("reading from file %s\n" % bam_filename)
samfile = pysam.Samfile(bam_filename, "rb")
for read in get_sam_iter(samfile, chrom):
count += 1
if count == 10000:
sys.stderr.write(".")
count = 0
add_read_count(read, chrom, ref_array, alt_array, other_array,
read_count_array, snp_index_array, snp_tab,
hap_tab, warned_pos, max_count, ind_idx)
# store results for this chromosome
ref_carray[:] = ref_array
alt_carray[:] = alt_array
other_carray[:] = other_array
read_count_carray[:] = read_count_array
sys.stderr.write("\n")
# write data to numpy arrays, so that they can be written to a txt
# file later
# columns are:
# chrom, pos, ref, alt, genotype, ref_count, alt_count, other_count
if args.txt_counts is not None:
write_txt_file(txt_counts, chrom, snp_tab, hap_tab, ind_idx,
ref_array, alt_array, other_array)
samfile.close()
if args.txt_counts:
# close the open txt file handler
txt_counts.close()
# check if any of the reads contained an unimplemented CIGAR
if unimplemented_CIGAR[0] > 0:
sys.stderr.write("WARNING: Encountered " +
str(unimplemented_CIGAR[0]) +
" instances of CIGAR codes: " +
str(unimplemented_CIGAR[1]) + ". Reads with these "
"CIGAR codes were skipped because they "
"are currently unimplemented.\n")
# set track statistics and close HDF5 files
sys.stderr.write("setting statistics for each chromosome\n")
for h5f in output_h5:
chromstat.set_stats(h5f, chrom_list)
h5f.close()
snp_tab_h5.close()
snp_index_h5.close()
if hap_h5:
hap_h5.close()
sys.stderr.write("done\n")
if (len(keep_cache) + len(discard_cache)) != 0:
sys.stderr.write("WARNING: failed to find pairs for %d "
"keep reads and %d discard reads on this "
"chromosome\n" %
(len(keep_cache), len(discard_cache)))
read_stats.discard_missing_pair += len(keep_cache) + len(discard_cache)
read_stats.write(sys.stderr)
if __name__ == "__main__":
sys.stderr.write("command line: %s\n" % " ".join(sys.argv))
sys.stderr.write("python version: %s\n" % sys.version)
sys.stderr.write("pysam version: %s\n" % pysam.__version__)
util.check_pysam_version()
parser = argparse.ArgumentParser()
parser.add_argument('input_bam',
help="input BAM or SAM file (must "
"be sorted!)")
parser.add_argument("output_bam",
help="output BAM or SAM file (not "
"sorted!)")
options = parser.parse_args()
main(options.input_bam, options.output_bam)
def main():
args = parse_args()
sys.stderr.write("command line: %s\n" % " ".join(sys.argv))
sys.stderr.write("python version: %s\n" % sys.version)
sys.stderr.write("pysam version: %s\n" % pysam.__version__)
sys.stderr.write("pytables version: %s\n" % tables.__version__)
util.check_pysam_version()
util.check_pytables_version()
# disable warnings that come from pytables when chromosome
# names are like 1, 2, 3 (instead of chr1, chr2, chr3)
warnings.filterwarnings('ignore', category=tables.NaturalNameWarning)
snp_tab_h5 = tables.open_file(args.snp_tab, "r")
snp_index_h5 = tables.open_file(args.snp_index, "r")
if args.haplotype:
hap_h5 = tables.open_file(args.haplotype, "r")
else:
hap_h5 = None
ref_count_h5 = tables.open_file(args.ref_as_counts, "w")
alt_count_h5 = tables.open_file(args.alt_as_counts, "w")
other_count_h5 = tables.open_file(args.other_as_counts, "w")
read_count_h5 = tables.open_file(args.read_counts, "w")
output_h5 = [ref_count_h5, alt_count_h5, other_count_h5, read_count_h5]
chrom_dict = {}
# initialize every chromosome in output files
chrom_list = chromosome.get_all_chromosomes(args.chrom)
for chrom in chrom_list:
for out_file in output_h5:
create_carray(out_file, chrom, args.data_type)
chrom_dict[chrom.name] = chrom
count = 0
dtype = None
if args.data_type == "uint8":
max_count = MAX_UINT8_COUNT
dtype = np.uint8
elif args.data_type == "uint16":
max_count = MAX_UINT16_COUNT
dtype = np.uint16
else:
raise NotImplementedError("unsupported datatype %s" % args.data_type)
# create a txt file to also holds the counts
if args.txt_counts is not None:
if os.path.splitext(args.txt_counts)[1] == ".gz":
txt_counts = gzip.open(args.txt_counts, 'wt+')
else:
txt_counts = open(args.txt_counts, 'w+')
for chrom in chrom_list:
sys.stderr.write("%s\n" % chrom.name)
if args.test_chrom:
if chrom.name != args.test_chrom:
sys.stderr.write("skipping because not test chrom\n")
continue
warned_pos = {}
# fetch SNP info for this chromosome
if chrom.name not in snp_tab_h5.root:
# no SNPs for this chromosome
sys.stderr.write("skipping %s because chromosome with this name "
"not found in SNP table\n" % chrom.name)
continue
sys.stderr.write("fetching SNPs\n")
snp_tab = snp_tab_h5.get_node("/%s" % chrom.name)
snp_index_array = snp_index_h5.get_node("/%s" % chrom.name)[:]
if hap_h5:
hap_tab = hap_h5.get_node("/%s" % chrom.name)
ind_dict, ind_idx = snptable.SNPTable().get_h5_sample_indices(
hap_h5, chrom, [args.individual])
if len(ind_idx) == 1:
ind_idx = ind_idx[0]
sys.stderr.write("index for individual %s is %d\n" %
(args.individual, ind_idx))
else:
raise ValueError("got sample indices for %d individuals, "
"but expected to get index for one "
"individual (%s)" % (len(ind_idx),
args.individual))
hap_tab = None
ind_idx = None
else:
hap_tab = None
ind_idx = None
# initialize count arrays for this chromosome to 0
ref_carray = get_carray(ref_count_h5, chrom)
alt_carray = get_carray(alt_count_h5, chrom)
other_carray = get_carray(other_count_h5, chrom)
read_count_carray = get_carray(read_count_h5, chrom)
ref_array = np.zeros(chrom.length, dtype)
alt_array = np.zeros(chrom.length, dtype)
other_array = np.zeros(chrom.length, dtype)
read_count_array = np.zeros(chrom.length, dtype)
# loop over all BAM files, pulling out reads
# for this chromosome
for bam_filename in args.bam_filenames:
sys.stderr.write("reading from file %s\n" % bam_filename)
samfile = pysam.Samfile(bam_filename, "rb")
for read in get_sam_iter(samfile, chrom):
count += 1
if count == 10000:
sys.stderr.write(".")
count = 0
add_read_count(read, chrom, ref_array, alt_array,
other_array, read_count_array,
snp_index_array, snp_tab, hap_tab,
warned_pos, max_count, ind_idx)
# store results for this chromosome
ref_carray[:] = ref_array
alt_carray[:] = alt_array
other_carray[:] = other_array
read_count_carray[:] = read_count_array
sys.stderr.write("\n")
# write data to numpy arrays, so that they can be written to a txt
# file later
# columns are:
# chrom, pos, ref, alt, genotype, ref_count, alt_count, other_count
if args.txt_counts is not None:
write_txt_file(txt_counts, chrom, snp_tab, hap_tab, ind_idx,
ref_array, alt_array, other_array)
samfile.close()
if args.txt_counts:
# close the open txt file handler
txt_counts.close()
# check if any of the reads contained an unimplemented CIGAR
if unimplemented_CIGAR[0] > 0:
sys.stderr.write("WARNING: Encountered " + str(unimplemented_CIGAR[0])
+ " instances of CIGAR codes: "
+ str(unimplemented_CIGAR[1]) + ". Reads with these "
"CIGAR codes were skipped because they "
"are currently unimplemented.\n")
# set track statistics and close HDF5 files
sys.stderr.write("setting statistics for each chromosome\n")
for h5f in output_h5:
chromstat.set_stats(h5f, chrom_list)
h5f.close()
snp_tab_h5.close()
snp_index_h5.close()
if hap_h5:
hap_h5.close()
sys.stderr.write("done\n")
def main():
args = parse_args()
sys.stderr.write("command line: %s\n" % " ".join(sys.argv))
sys.stderr.write("python version: %s\n" % sys.version)
sys.stderr.write("pysam version: %s\n" % pysam.__version__)
sys.stderr.write("pytables version: %s\n" % tables.__version__)
util.check_pysam_version()
util.check_pytables_version()
snp_tab_h5 = tables.open_file(args.snp_tab, "r")
snp_index_h5 = tables.open_file(args.snp_index, "r")
if args.haplotype:
hap_h5 = tables.open_file(args.haplotype, "r")
ind_idx = lookup_individual_index(args.samples, args.individual)
else:
hap_h5 = None
ind_idx = None
ref_count_h5 = tables.open_file(args.ref_as_counts, "w")
alt_count_h5 = tables.open_file(args.alt_as_counts, "w")
other_count_h5 = tables.open_file(args.other_as_counts, "w")
read_count_h5 = tables.open_file(args.read_counts, "w")
output_h5 = [ref_count_h5, alt_count_h5, other_count_h5, read_count_h5]
chrom_dict = {}
# initialize every chromosome in output files
chrom_list = chromosome.get_all_chromosomes(args.chrom)
for chrom in chrom_list:
for out_file in output_h5:
create_carray(out_file, chrom, args.data_type)
chrom_dict[chrom.name] = chrom
count = 0
dtype = None
if args.data_type == "uint8":
max_count = MAX_UINT8_COUNT
dtype = np.uint8
elif args.data_type == "uint16":
max_count = MAX_UINT16_COUNT
dtype = np.uint16
else:
raise NotImplementedError("unsupported datatype %s" % args.data_type)
# create a list to hold the counts that will be later written
# to a txt file
if args.text_counts is not None:
txt_counts = list()
for chrom in chrom_list:
sys.stderr.write("%s\n" % chrom.name)
warned_pos = {}
# fetch SNP info for this chromosome
if chrom.name not in snp_tab_h5.root:
# no SNPs for this chromosome
continue
sys.stderr.write("fetching SNPs\n")
snp_tab = snp_tab_h5.get_node("/%s" % chrom.name)
snp_index_array = snp_index_h5.get_node("/%s" % chrom.name)[:]
if hap_h5:
hap_tab = hap_h5.get_node("/%s" % chrom.name)
else:
hap_tab = None
# initialize count arrays for this chromosome to 0
ref_carray = get_carray(ref_count_h5, chrom)
alt_carray = get_carray(alt_count_h5, chrom)
other_carray = get_carray(other_count_h5, chrom)
read_count_carray = get_carray(read_count_h5, chrom)
ref_array = np.zeros(chrom.length, dtype)
alt_array = np.zeros(chrom.length, dtype)
other_array = np.zeros(chrom.length, dtype)
read_count_array = np.zeros(chrom.length, dtype)
# loop over all BAM files, pulling out reads
# for this chromosome
for bam_filename in args.bam_filenames:
sys.stderr.write("reading from file %s\n" % bam_filename)
samfile = pysam.Samfile(bam_filename, "rb")
for read in get_sam_iter(samfile, chrom):
count += 1
if count == 10000:
sys.stderr.write(".")
count = 0
add_read_count(read, chrom, ref_array, alt_array,
other_array, read_count_array,
snp_index_array, snp_tab, hap_tab,
warned_pos, max_count, ind_idx)
# store results for this chromosome
ref_carray[:] = ref_array
alt_carray[:] = alt_array
other_carray[:] = other_array
read_count_carray[:] = read_count_array
sys.stderr.write("\n")
# write data to numpy arrays, so that they can be written to a txt
# file later
# columns are:
# chrom, pos, ref, alt, genotype, ref_count, alt_count, other_count
if args.text_counts is not None:
chrom = np.tile(chrom.name, len(snp_tab))
pos = np.array([snp['pos'] for snp in snp_tab])
ref = np.array([snp['allele1'] for snp in snp_tab])
alt = np.array([snp['allele2'] for snp in snp_tab])
if hap_tab is not None:
genotype = np.array([str(hap[0])+"|"+str(hap[1])
for hap in hap_tab])
else:
genotype = np.empty((len(snp_tab), 0))
txt_counts.append(
np.column_stack((chrom, pos, ref, alt, genotype,
ref_array[pos-1],
alt_array[pos-1],
other_array[pos-1]))
)
samfile.close()
# write the txt_counts np arrays to a txt file
if args.text_counts is not None:
# we use vstack to combine np arrays row-wise into a multi-dimensional
# array
np.savetxt(args.text_counts, np.vstack(tuple(txt_counts)),
fmt="%1s", delimiter=" ")
# set track statistics and close HDF5 files
sys.stderr.write("setting statistics for each chromosome\n")
for h5f in output_h5:
chromstat.set_stats(h5f, chrom_list)
h5f.close()
snp_tab_h5.close()
snp_index_h5.close()
if hap_h5:
hap_h5.close()
sys.stderr.write("done\n")
def main():
args = parse_args()
sys.stderr.write("command line: %s\n" % " ".join(sys.argv))
sys.stderr.write("python version: %s\n" % sys.version)
sys.stderr.write("pysam version: %s\n" % pysam.__version__)
sys.stderr.write("pytables version: %s\n" % tables.__version__)
util.check_pysam_version()
util.check_pytables_version()
snp_tab_h5 = tables.open_file(args.snp_tab, "r")
snp_index_h5 = tables.open_file(args.snp_index, "r")
if args.haplotype:
hap_h5 = tables.open_file(args.haplotype, "r")
else:
hap_h5 = None
ref_count_h5 = tables.open_file(args.ref_as_counts, "w")
alt_count_h5 = tables.open_file(args.alt_as_counts, "w")
other_count_h5 = tables.open_file(args.other_as_counts, "w")
read_count_h5 = tables.open_file(args.read_counts, "w")
output_h5 = [ref_count_h5, alt_count_h5, other_count_h5, read_count_h5]
chrom_dict = {}
# initialize every chromosome in output files
chrom_list = chromosome.get_all_chromosomes(args.chrom)
for chrom in chrom_list:
for out_file in output_h5:
create_carray(out_file, chrom, args.data_type)
chrom_dict[chrom.name] = chrom
count = 0
dtype = None
if args.data_type == "uint8":
max_count = MAX_UINT8_COUNT
dtype = np.uint8
elif args.data_type == "uint16":
max_count = MAX_UINT16_COUNT
dtype = np.uint16
else:
raise NotImplementedError("unsupported datatype %s" % args.data_type)
# create a txt file to also holds the counts
if args.txt_counts is not None:
if os.path.splitext(args.txt_counts)[1] == ".gz":
txt_counts = gzip.open(args.txt_counts, 'a+')
else:
txt_counts = open(args.txt_counts, 'a+')
for chrom in chrom_list:
sys.stderr.write("%s\n" % chrom.name)
warned_pos = {}
# fetch SNP info for this chromosome
if chrom.name not in snp_tab_h5.root:
# no SNPs for this chromosome
continue
sys.stderr.write("fetching SNPs\n")
snp_tab = snp_tab_h5.get_node("/%s" % chrom.name)
snp_index_array = snp_index_h5.get_node("/%s" % chrom.name)[:]
if hap_h5:
hap_tab = hap_h5.get_node("/%s" % chrom.name)
ind_idx = snptable.SNPTable().get_h5_sample_indices(
hap_h5, chrom, [args.individual])[1]
if len(ind_idx) != 0:
ind_idx = ind_idx[0]
else:
hap_tab = None
ind_idx = None
else:
hap_tab = None
ind_idx = None
# initialize count arrays for this chromosome to 0
ref_carray = get_carray(ref_count_h5, chrom)
alt_carray = get_carray(alt_count_h5, chrom)
other_carray = get_carray(other_count_h5, chrom)
read_count_carray = get_carray(read_count_h5, chrom)
ref_array = np.zeros(chrom.length, dtype)
alt_array = np.zeros(chrom.length, dtype)
other_array = np.zeros(chrom.length, dtype)
read_count_array = np.zeros(chrom.length, dtype)
# loop over all BAM files, pulling out reads
# for this chromosome
for bam_filename in args.bam_filenames:
sys.stderr.write("reading from file %s\n" % bam_filename)
samfile = pysam.Samfile(bam_filename, "rb")
for read in get_sam_iter(samfile, chrom):
count += 1
if count == 10000:
sys.stderr.write(".")
count = 0
add_read_count(read, chrom, ref_array, alt_array, other_array,
read_count_array, snp_index_array, snp_tab,
hap_tab, warned_pos, max_count, ind_idx)
# store results for this chromosome
ref_carray[:] = ref_array
alt_carray[:] = alt_array
other_carray[:] = other_array
read_count_carray[:] = read_count_array
sys.stderr.write("\n")
# write data to numpy arrays, so that they can be written to a txt
# file later
# columns are:
# chrom, pos, ref, alt, genotype, ref_count, alt_count, other_count
if args.txt_counts is not None:
chrom = np.tile(chrom.name, len(snp_tab))
pos = np.array([snp['pos'] for snp in snp_tab])
ref = np.array([snp['allele1'] for snp in snp_tab])
alt = np.array([snp['allele2'] for snp in snp_tab])
if hap_tab is not None:
genotype = np.array(
[str(hap[0]) + "|" + str(hap[1]) for hap in hap_tab])
else:
genotype = np.empty((len(snp_tab), 0))
# write an np array to a txt file
np.savetxt(txt_counts,
np.column_stack(
(chrom, pos, ref, alt, genotype,
ref_array[pos - 1], alt_array[pos - 1],
other_array[pos - 1])),
fmt="%1s",
delimiter=" ")
samfile.close()
if args.txt_counts:
# close the open txt file handler
txt_counts.close()
# check if any of the reads contained an unimplemented CIGAR
sys.stderr.write(
"WARNING: Encountered " + str(unimplemented_CIGAR[0]) +
" instances of any of the following CIGAR codes: " +
str(unimplemented_CIGAR[1]) +
". The regions of reads with these CIGAR codes were skipped because these CIGAR codes are currently unimplemented.\n"
)
# set track statistics and close HDF5 files
sys.stderr.write("setting statistics for each chromosome\n")
for h5f in output_h5:
chromstat.set_stats(h5f, chrom_list)
h5f.close()
snp_tab_h5.close()
snp_index_h5.close()
if hap_h5:
hap_h5.close()
sys.stderr.write("done\n")
haplotype_filename=haplotype_filename)
filter_reads(files, max_seqs=max_seqs, max_snps=max_snps,
samples=samples)
files.close()
if __name__ == '__main__':
sys.stderr.write("command line: %s\n" % " ".join(sys.argv))
sys.stderr.write("python version: %s\n" % sys.version)
sys.stderr.write("pysam version: %s\n" % pysam.__version__)
sys.stderr.write("pytables version: %s\n" % tables.__version__)
util.check_pysam_version()
util.check_pytables_version()
util.check_python_version()
options = parse_options()
samples = parse_samples(options.samples)
main(options.bam_filename,
is_paired_end=options.is_paired_end, is_sorted=options.is_sorted,
max_seqs=options.max_seqs, max_snps=options.max_snps,
output_dir=options.output_dir,
snp_dir=options.snp_dir,
snp_tab_filename=options.snp_tab,
snp_index_filename=options.snp_index,
haplotype_filename=options.haplotype,
samples=samples)
