def visualize_sv_calls(args, dbo_args, endpoint_args):
in_svcalls_bedpe_file = args.filter_bedpe_file
bcd13_file = dbo_args.bcd13_file
bcd21_file = endpoint_args.bcd21_file
faidx_file = args.faidx_file
cal_2d_overlapping_barcodes_binary = args.cal_2d_overlapping_barcodes
cal_read_depth_from_bcd21_binary = args.cal_read_depth_from_bcd21
out_dir = args.image_out_dir
out_prefix = args.bam_name
my_utils.make_dir(out_dir)
chr_len_list = my_utils.get_chr_length(faidx_file)
tid2chrname_list, chrname2tid_dict = my_utils.get_chrnames(faidx_file)
in_svcall_list = read_svcall_bedpe_file(in_svcalls_bedpe_file, chrname2tid_dict)
for i in range(0, len(in_svcall_list)):
in_svcall_list[i].format()
plot_depth(cal_read_depth_from_bcd21_binary, bcd21_file, in_svcall_list, faidx_file, chr_len_list, tid2chrname_list, chrname2tid_dict, out_dir, out_prefix)
plot_twin_window_barcode_similarity(in_svcall_list, bcd13_file, faidx_file, out_dir, chr_len_list, tid2chrname_list, chrname2tid_dict, out_prefix)
flank_dist = 100 * 1000 # set flank distance to be 100 kb
plot_heatmap (in_svcall_list, bcd21_file, faidx_file, out_dir, flank_dist, chr_len_list, tid2chrname_list, chrname2tid_dict, cal_2d_overlapping_barcodes_binary, out_prefix)
return
def plot_heatmap(in_svcall_list, bcd21_file, faidx_file, out_dir, flank_dist, chr_len_list, tid2chrname_list, chrname2tid_dict, cal_2d_overlapping_barcodes_binary, out_prefix):
if os.path.exists(cal_2d_overlapping_barcodes_binary) == False:
my_utils.myprint('ERROR! The binary file doesn\'t exist: %s' % cal_2d_overlapping_barcodes_binary)
my_utils.myprint('Skipped plotting the heat maps')
return
out_dir = os.path.join(out_dir, '2D_heatmap')
my_utils.make_dir(out_dir)
my_utils.myprint('plotting heat maps of overlapping barcodes')
target_region_bedpe_list = generate_target_region_bedpe_list(in_svcall_list, chr_len_list, flank_dist, chrname2tid_dict)
target_region_bedpe_file = os.path.join(out_dir, 'target_region.bedpe')
target_region_bedpe_fp = my_utils.gzopen(target_region_bedpe_file, 'w')
for bedpe1 in target_region_bedpe_list:
target_region_bedpe_fp.write(bedpe1.output_svcall() + endl)
target_region_bedpe_fp.close()
target_region_2d_ovl_with_low_mapq_file = os.path.join(out_dir, '%s.2d_heatmap.with_low_mapq_reads.txt' % out_prefix)
bin_size = 1000
max_ovl_num = 100
cmd_args_list1 = [cal_2d_overlapping_barcodes_binary, bcd21_file, target_region_bedpe_file, target_region_2d_ovl_with_low_mapq_file, faidx_file, str(bin_size), '1']
subprocess.call(cmd_args_list1)
plot_2d_barcodes.plot_2d_overlapping_barcodes(target_region_2d_ovl_with_low_mapq_file, target_region_bedpe_list, bin_size, max_ovl_num, out_dir, out_prefix)
return
def plot_2d_overlapping_barcodes(in_2d_values_file, target_region_bedpe_list,
bin_size, max_ovl_num, out_dir, out_prefix):
my_utils.make_dir(out_dir)
in_2d_values_fp = my_utils.gzopen(in_2d_values_file, 'r')
region_title = ''
xmin, xmax, ymin, ymax = (0, 0, 0, 0)
xsize = 0
ysize = 0
current_x = 0
current_y = 0
ovl_2d_array = np.zeros(shape=(xsize, ysize), dtype=np.int32)
while 1:
line = in_2d_values_fp.readline()
if not line: break
line = line.strip()
if line[0:2] == '##':
if (xsize and ysize):
plot_one_bedpe(out_dir, target_region_bedpe_list, out_prefix,
region_title, ovl_2d_array, xmin, xmax, ymin,
ymax, bin_size)
region_title = line[2:]
xmin, xmax, ymin, ymax = (0, 0, 0, 0)
xsize = 0
ysize = 0
current_x = 0
current_y = 0
continue
if line[0:5] == '#xmin':
line = line.split('=')
xmin, xmax, ymin, ymax = line[1].split(',')
xmin = int(xmin)
xmax = int(xmax)
ymin = int(ymin)
ymax = int(ymax)
xsize = int((xmax - xmin) / bin_size)
ysize = int((ymax - ymin) / bin_size)
ovl_2d_array = np.zeros(shape=(xsize, ysize), dtype=np.int32)
current_x = 0
current_y = 0
continue
line = line.split(tab)
for current_y in range(0, len(line)):
ovl_2d_array[current_x][current_y] = int(line[current_y])
if ovl_2d_array[current_x][current_y] > max_ovl_num:
ovl_2d_array[current_x][current_y] = max_ovl_num
current_x += 1
if (xsize and ysize):
plot_one_bedpe(out_dir, target_region_bedpe_list, out_prefix,
region_title, ovl_2d_array, xmin, xmax, ymin, ymax,
bin_size)
in_2d_values_fp.close()
return
def plot_twin_window_barcode_similarity(in_svcall_list, bcd13_file, faidx_file, out_dir, chr_len_list, tid2chrname_list, chrname2tid_dict, out_prefix):
out_dir = os.path.join(out_dir, 'twin_window_barcode_similarity')
my_utils.make_dir(out_dir)
wg_pvalue_list, bin_size = get_wg_pvalue_list_from_bcd13_file(bcd13_file, chr_len_list)
max_length_in_one_figure = 500 * 1000
for svcall in in_svcall_list:
if svcall.svtype == 'DEL' or svcall.svtype == 'DUP': continue
flank_dist = abs(svcall.key2 - svcall.key1)
if flank_dist > 50 * 1000: flank_dist = 50 * 1000
if flank_dist < 10 * 1000: flank_dist = 10 * 1000
if abs(svcall.key2 - svcall.key1) > max_length_in_one_figure:
out_file1 = os.path.join(out_dir, '%s.%s.breakpoint1.twin_window_barcode_similarity.png' % (out_prefix, svcall.sv_id ))
reg_start = svcall.start1 - flank_dist
reg_end = svcall.end1 + flank_dist
if reg_start < 0: reg_start = 0
if reg_end > chr_len_list[svcall.tid1]: reg_end = chr_len_list[svcall.tid1]
bk_pos1 = svcall.start1
bk_pos2 = bk_pos1
figure_title = '%s, %s (breakpoint 1)' % (svcall.sv_id, svcall.svtype)
plot_twin_window_barcode_similarity_for1region(svcall.chrm1, svcall.tid1, reg_start, reg_end, bk_pos1, bk_pos2, out_file1, figure_title, wg_pvalue_list, bin_size)
out_file2 = os.path.join(out_dir, '%s.%s.breakpoint2.twin_window_barcode_similarity.png' % (out_prefix, svcall.sv_id ))
reg_start = svcall.start2 - flank_dist
reg_end = svcall.end2 + flank_dist
bk_pos1 = svcall.start2
bk_pos2 = bk_pos1
if reg_start < 0: reg_start = 0
if reg_end > chr_len_list[svcall.tid2]: reg_end = chr_len_list[svcall.tid2]
figure_title = '%s, %s (breakpoint 2)' % (svcall.sv_id, svcall.svtype)
plot_twin_window_barcode_similarity_for1region(svcall.chrm2, svcall.tid2, reg_start, reg_end, bk_pos1, bk_pos2, out_file2, figure_title, wg_pvalue_list, bin_size)
else:
out_file = os.path.join(out_dir, '%s.%s.both_breakpoints.twin_window_barcode_similarity.png' % (out_prefix, svcall.sv_id))
reg_start = svcall.start1 - flank_dist
reg_end = svcall.end2 + flank_dist
if reg_start < 0: reg_start = 0
if reg_end > chr_len_list[svcall.tid1]: reg_end = chr_len_list[svcall.tid1]
bk_pos1 = svcall.start1
bk_pos2 = svcall.start2
figure_title = '%s, %s (breakpoint 1 and 2)' % (svcall.sv_id, svcall.svtype)
plot_twin_window_barcode_similarity_for1region(svcall.chrm1, svcall.tid1, reg_start, reg_end, bk_pos1, bk_pos2, out_file, figure_title, wg_pvalue_list, bin_size)
return
def plot_depth(cal_read_depth_from_bcd21_binary, bcd21_file, in_svcalls_list, faidx_file, chr_len_list, tid2chrname_list, chrname2tid_dict, out_dir, out_prefix):
if os.path.exists(cal_read_depth_from_bcd21_binary) == False:
my_utils.myprint('ERROR! The binary file doesn\'t exist:%s\Failed to plot read depth' % cal_read_depth_from_bcd21_binary)
return
if os.path.exists(bcd21_file) == False:
my_utils.myprint('ERROR! The bcd21 file doesn\'t exist:%s\Failed to plot read depth' % bcd21_file)
return
if os.path.exists(faidx_file) == False:
my_utils.myprint('ERROR! The fasta index file doesn\'t exist:%s\Failed to plot read depth' % faidx_file)
return
out_dir = os.path.join(out_dir, 'read_depth')
my_utils.make_dir(out_dir)
bin_size = 500
read_depth_file = os.path.join(out_dir, '%s.read_depth.txt' % out_prefix)
cmd_args_list = [cal_read_depth_from_bcd21_binary, bcd21_file, read_depth_file, faidx_file, str(bin_size), '20']
my_utils.myprint('calculating read depth from file: %s' % bcd21_file)
subprocess.call(cmd_args_list)
my_utils.myprint('plotting read depth')
wg_high_mapq_depth_list, wg_total_depth_list, bin_size = plot_read_depth.get_wg_depth_list(read_depth_file, chr_len_list)
wg_total_depth = 0
wg_n_bin = 0
for tid in range(0, len(wg_high_mapq_depth_list)):
for depth in wg_high_mapq_depth_list[tid]:
wg_total_depth += depth
wg_n_bin += 1
wg_avg_depth = float(wg_total_depth) / wg_n_bin
for svcall in in_svcalls_list:
if svcall.chrm1 != svcall.chrm2: continue
out_file = os.path.join(out_dir, '%s.%s.read_depth.png' % (out_prefix, svcall.sv_id))
figure_title = 'Read depth (%s, %d bp %s)' % (svcall.sv_id, svcall.end2 - svcall.start1, svcall.svtype)
plot_read_depth.plot_read_depth_for1region(svcall.chrm1, svcall.tid1, svcall.start1, svcall.end2, out_file, figure_title, wg_high_mapq_depth_list, wg_total_depth_list, chr_len_list, bin_size, wg_avg_depth)
os.remove(read_depth_file)
return