def genotype_DTS_regions(dts_list_file, regions_file, contigs, window):
"""
Get copy number from each DenseTrackSet for each region in the given file.
"""
# Load supported contigs.
with open(contigs, "r") as fh:
supported_contigs = set([line.strip().split("\t")[0] for line in fh])
regions_by_chromosome = get_regions(regions_file, supported_contigs)
# Load DTS file list.
with open(dts_list_file, "r") as fh:
dts_list = [line.strip() for line in fh]
# Build a sorted list of sample names.
dts_list.sort()
sample_names = [os.path.basename(dts) for dts in dts_list]
yield ["chromosome", "start", "end"] + sample_names
for chromosome, regions in regions_by_chromosome.iteritems():
copies_by_sample = {}
for i in xrange(len(dts_list)):
dts = dts_list[i]
sample_name = sample_names[i]
sample = wnd_cp_indiv(dts, contigs, window)
copies_by_sample[sample_name] = sample.get_cp_by_regions(
chromosome,
regions["starts"],
regions["ends"]
)
for i in xrange(len(regions["starts"])):
yield list([chromosome, str(regions["starts"][i]), str(regions["ends"][i])] +
[str(copies_by_sample[sample][i]) for sample in sample_names])
parser.add_argument('--window_size', type=int, default=None, help='Size of SUNK/wssd sliding windows')
parser.add_argument('--min_ref_cp_delta', dest='min_d', type=float, default=0, help='Smallest difference in cp between ref and sample to consider (Default: %(default)s)')
parser.add_argument('--no_P_value_adjust', dest='P_adjust', action='store_false')
parser.add_argument('--min_mu', default=0.5, type=float, help='Minimum cluster mean (Default: %(default)s)')
parser.add_argument('--subset_indivs', default=None, help='Colon-separated list of individuals to consider (Default: %(default)s)')
o = parser.parse_args()
subset_indivs = o.subset_indivs
if subset_indivs != None:
subset_indivs = subset_indivs.split(":")
subset_indivs = list(set(subset_indivs))
indiv_DTS = wnd_cp_indiv(o.fn_indiv_DTS, o.fn_contigs, o.window_size)
indiv_id = o.fn_indiv_DTS.split("/")[-1].replace("500_bp_","")
ref_DTSs = {}
dCGHs = {}
for fn_ref in o.fn_ref_DTS.split(":"):
dCGHs[fn_ref.split("/")[-1].replace("500_bp_","")] = dCGH(o.fn_indiv_DTS,
fn_ref,
o.fn_contigs,
o.window_size)
ref_DTSs[fn_ref.split("/")[-1].replace("500_bp_","")] = wnd_cp_indiv(fn_ref,
o.fn_contigs,
o.window_size)
call_table = cluster.indiv_callset_table(o.fn_call_table)
opts.add_option('','--fn_DTS',dest='fn_DTS', default=None)
opts.add_option('','--contigs',dest='fn_contigs', default=None)
opts.add_option('','--wnd_size',dest='wnd_size', type=int, default=None)
#opts.add_option('','--wnd_slide',dest='wnd_slide', type=int, default=None)
opts.add_option('','--out_dir',dest='out_dir')
opts.add_option('','--fn_out',dest='fn_out')
opts.add_option('','--contig_prefix',dest='contig_prefix', default="")
opts.add_option('','--DTS_prefix',dest='DTS_prefix', default="500_bp_")
opts.add_option('','--output_contigs',dest='output_contigs', default="/net/eichler/vol7/home/psudmant/genomes/contigs/hg19_contigs.txt")
(o, args) = opts.parse_args()
#usage, init, then run
indiv = o.fn_DTS.split("/")[-1].replace("500_bp_","")
wnd_cp = wnd_cp_indiv(o.fn_DTS, o.fn_contigs, o.wnd_size)
"""
outstr:
chr start end indiv 0 0 0 color
"""
c_out = output(o.contig_prefix, o.output_contigs)
for contig in wnd_cp.contigs:
print >>stderr, contig
cps = wnd_cp.get_cps_by_chr(contig)
wnd_starts, wnd_ends = wnd_cp.get_wnds_by_chr(contig)
prev_start = 0
for i in xrange(0, cps.shape[0]-1):
s, e = wnd_starts[i], wnd_ends[i]
def init_from_DTS(cls, **kwargs):
"""
requires the below inputs
gglob.init_from_DTS(DTS_dir = DTS_dir,
DTS_prefix = DTS_prefix,
sunk_DTS_dir = sunk_DTS_dir,
sunk_DTS_prefix = sunk_DTS_prefix,
wnd_size = wnd_size,
indivs = indivs,
contig = contig,
fn_contigs = fn_contigs,
fn_sunk_contigs = fn_sunk_contigs)
"""
DTS_dir = kwargs["DTS_dir"]
DTS_prefix = kwargs["DTS_prefix"]
sunk_DTS_dir = kwargs["sunk_DTS_dir"]
sunk_DTS_prefix = kwargs["sunk_DTS_prefix"]
wnd_size = kwargs['wnd_size']
wnd_slide = kwargs['wnd_slide']
indivs = kwargs['indivs']
contig = kwargs['contig']
fn_contigs = kwargs['fn_contigs']
fn_sunk_contigs = kwargs['fn_sunk_contigs']
DTS_pre = "%s/%s" % (DTS_dir, DTS_prefix)
sunk_DTS_pre = "%s/%s" % (sunk_DTS_dir, DTS_prefix)
n_indivs = len(indivs)
t = time.time()
rand_wnd_cp = wnd_cp_indiv("%s%s" % (DTS_pre, indivs[0]), fn_contigs,
wnd_size)
wnd_starts, wnd_ends = rand_wnd_cp.get_wnds_by_chr(contig)
cp_matrix = np.zeros((n_indivs, wnd_starts.shape[0]))
rand_sunk_wnd_cp = wnd_cp_indiv("%s%s" % (sunk_DTS_pre, indivs[0]),
fn_sunk_contigs, wnd_size)
sunk_wnd_starts, sunk_wnd_ends = rand_sunk_wnd_cp.get_wnds_by_chr(
contig)
sunk_cp_matrix = np.zeros((n_indivs, sunk_wnd_starts.shape[0]))
correct = not (contig in ["chrY", "chrX"])
for i, indiv in enumerate(indivs):
print(indiv, file=stderr)
wnd_cp = wnd_cp_indiv("%s%s" % (DTS_pre, indiv), fn_contigs,
wnd_size)
cp_matrix[i, :] = wnd_cp.get_cps_by_chr(contig, correct=correct)
sunk_wnd_cp = wnd_cp_indiv("%s%s" % (sunk_DTS_pre, indiv),
fn_sunk_contigs, wnd_size)
sunk_cp_matrix[i, :] = sunk_wnd_cp.get_cps_by_chr(contig,
correct=correct)
return cls(indivs=indivs,
wnd_size=wnd_size,
wnd_slide=wnd_slide,
contig=contig,
wnd_starts=wnd_starts,
wnd_ends=wnd_ends,
cp_matrix=cp_matrix,
sunk_wnd_starts=sunk_wnd_starts,
sunk_wnd_ends=sunk_wnd_ends,
sunk_cp_matrix=sunk_cp_matrix)
def init_from_DTS(cls, **kwargs):
"""
requires the below inputs
gglob.init_from_DTS(DTS_dir = DTS_dir,
DTS_prefix = DTS_prefix,
sunk_DTS_dir = sunk_DTS_dir,
sunk_DTS_prefix = sunk_DTS_prefix,
wnd_size = wnd_size,
indivs = indivs,
contig = contig,
fn_contigs = fn_contigs,
fn_sunk_contigs = fn_sunk_contigs)
"""
DTS_dir = kwargs["DTS_dir"]
DTS_prefix = kwargs["DTS_prefix"]
sunk_DTS_dir = kwargs["sunk_DTS_dir"]
sunk_DTS_prefix = kwargs["sunk_DTS_prefix"]
wnd_size = kwargs['wnd_size']
wnd_slide = kwargs['wnd_slide']
indivs = kwargs['indivs']
contig = kwargs['contig']
fn_contigs = kwargs['fn_contigs']
fn_sunk_contigs = kwargs['fn_sunk_contigs']
DTS_pre="%s/%s"%(DTS_dir, DTS_prefix)
sunk_DTS_pre="%s/%s"%(sunk_DTS_dir, DTS_prefix)
n_indivs = len(indivs)
t = time.time()
rand_wnd_cp = wnd_cp_indiv("%s%s"%(DTS_pre, indivs[0]), fn_contigs, wnd_size)
wnd_starts, wnd_ends = rand_wnd_cp.get_wnds_by_chr(contig)
cp_matrix = np.zeros((n_indivs, wnd_starts.shape[0]))
rand_sunk_wnd_cp = wnd_cp_indiv("%s%s"%(sunk_DTS_pre, indivs[0]), fn_sunk_contigs, wnd_size)
sunk_wnd_starts, sunk_wnd_ends = rand_sunk_wnd_cp.get_wnds_by_chr(contig)
sunk_cp_matrix = np.zeros((n_indivs, sunk_wnd_starts.shape[0]))
correct = not (contig in ["chrY", "chrX"])
for i, indiv in enumerate(indivs):
print >> stderr, indiv
wnd_cp = wnd_cp_indiv("%s%s"%(DTS_pre, indiv),
fn_contigs,
wnd_size)
cp_matrix[i,:] = wnd_cp.get_cps_by_chr(contig, correct=correct)
sunk_wnd_cp = wnd_cp_indiv("%s%s"%(sunk_DTS_pre, indiv),
fn_sunk_contigs,
wnd_size)
sunk_cp_matrix[i,:] = sunk_wnd_cp.get_cps_by_chr(contig, correct=correct)
return cls(indivs = indivs,
wnd_size = wnd_size,
wnd_slide = wnd_slide,
contig = contig,
wnd_starts = wnd_starts,
wnd_ends = wnd_ends,
cp_matrix = cp_matrix,
sunk_wnd_starts = sunk_wnd_starts,
sunk_wnd_ends = sunk_wnd_ends,
sunk_cp_matrix = sunk_cp_matrix)
opts.add_option('', '--fn_out', dest='fn_out')
opts.add_option('', '--contig_prefix', dest='contig_prefix', default="")
opts.add_option('', '--DTS_prefix', dest='DTS_prefix', default="500_bp_")
opts.add_option(
'',
'--output_contigs',
dest='output_contigs',
default=
"/net/eichler/vol27/projects/human_diversity/nobackups/hsiehph/genomicData/EEElab/read_depth_genotyper/hg19_contigs.txt"
)
(o, args) = opts.parse_args()
#usage, init, then run
indiv = o.fn_DTS.split("/")[-1].replace("500_bp_", "")
wnd_cp = wnd_cp_indiv(o.fn_DTS, o.fn_contigs, o.wnd_size)
"""
outstr:
chr start end indiv 0 0 0 color
"""
c_out = output(o.contig_prefix, o.output_contigs)
for contig in wnd_cp.contigs:
print(contig, file=stderr)
cps = wnd_cp.get_cps_by_chr(contig)
wnd_starts, wnd_ends = wnd_cp.get_wnds_by_chr(contig)
prev_start = 0
for i in range(0, cps.shape[0] - 1):
s, e = wnd_starts[i], wnd_ends[i]
chrs = []
for c in tbx_gc.contigs:
if (not "random" in c) and (not "X" in c) and (not "Y" in c):
chrs.append(c)
fn_contigs = o.fn_contigs
wnd = int(o.window_size)
##
#cutoff_scale=float(bp_cutoff_scale/wnd)
#max_merge = o.max_merge_dif
#cp_data = dCGH(o.fn_in_DTS,o.fn_ref_DTS,o.fn_contigs,wnd)
#segment_callset = callset()
#caller_by_chr = {}
#dCGH(o.fn_in_DTS,o.fn_ref_DTS,o.fn_contigs,wnd)
cp_data = wnd_cp_indiv(o.fn_in_DTS,o.fn_contigs,wnd)
GC_DTS = DenseTrackSet(o.fn_contigs,o.fn_GC_DTS,overwrite=False,openMode='r')
null_dist = null_distribution(tbx_gc)
for chr in chrs:
print >>stderr,"%s..."%chr
magnitude_vect = cp_data.get_cps_by_chr(chr)
starts_vect,ends_vect = cp_data.get_wnds_by_chr(chr)
#plot_GC(chr,tbx_gc,magnitude_vect,starts_vect,ends_vect)
#print magnitude_vect[0:1000]
gapped_wnds = cp_data.get_overlapping_wnds(chr,tbx_gaps)
segdup_wnds = cp_data.get_overlapping_wnds(chr,tbx_dups)
null_dist.add(magnitude_vect,[gapped_wnds,segdup_wnds])
parser.add_argument(
'--subset_indivs',
default=None,
help=
'Colon-separated list of individuals to consider (Default: %(default)s)'
)
o = parser.parse_args()
subset_indivs = o.subset_indivs
if subset_indivs != None:
subset_indivs = subset_indivs.split(":")
subset_indivs = list(set(subset_indivs))
indiv_DTS = wnd_cp_indiv(o.fn_indiv_DTS, o.fn_contigs, o.window_size)
indiv_id = o.fn_indiv_DTS.split("/")[-1].replace("500_bp_", "")
ref_DTSs = {}
dCGHs = {}
for fn_ref in o.fn_ref_DTS.split(":"):
dCGHs[fn_ref.split("/")[-1].replace("500_bp_", "")] = dCGH(
o.fn_indiv_DTS, fn_ref, o.fn_contigs, o.window_size)
ref_DTSs[fn_ref.split("/")[-1].replace("500_bp_", "")] = wnd_cp_indiv(
fn_ref, o.fn_contigs, o.window_size)
call_table = cluster.indiv_callset_table(o.fn_call_table)
if o.limit_to_chr:
call_table.filter_by_chr(o.limit_to_chr)