def sanity_check_lifted_nagalakshmi_file(fn):
gtf_fn = '/home/jah/projects/arlen/data/organisms/saccharomyces_cerevisiae/EF4/transcriptome/genes.gtf'
CDSs = gtf.get_CDSs(gtf_fn)
gtf_dict = {t.name: t for t in CDSs}
genes = read_nagalakshmi_file(fn)
discrepancies_after = defaultdict(list)
for name in genes:
if name not in gtf_dict:
print name, 'not in gtf_dict'
continue
start, end = genes[name]['SGD_Start'], genes[name]['SGD_End']
chrom = genes[name]['Chrom']
if start > end:
start, end = end, start
end -= 1
if start != gtf_dict[name].start or end != gtf_dict[name].end:
#print name, chrom
#print start, end
#print gtf_dict[name].start, gtf_dict[name].end
#raw_input()
discrepancies_after[chrom].append(name)
return discrepancies_after
def sanity_check_lifted_nagalakshmi_file(fn):
gtf_fn = '/home/jah/projects/arlen/data/organisms/saccharomyces_cerevisiae/EF4/transcriptome/genes.gtf'
CDSs = gtf.get_CDSs(gtf_fn)
gtf_dict = {t.name: t for t in CDSs}
genes = read_nagalakshmi_file(fn)
discrepancies_after = defaultdict(list)
for name in genes:
if name not in gtf_dict:
print name, 'not in gtf_dict'
continue
start, end = genes[name]['SGD_Start'], genes[name]['SGD_End']
chrom = genes[name]['Chrom']
if start > end:
start, end = end, start
end -= 1
if start != gtf_dict[name].start or end != gtf_dict[name].end:
#print name, chrom
#print start, end
#print gtf_dict[name].start, gtf_dict[name].end
#raw_input()
discrepancies_after[chrom].append(name)
return discrepancies_after
def sanity_check_nagalakshmi_file():
gtf_fn = '/home/jah/projects/arlen/data/organisms/saccharomyces_cerevisiae/SGD1.01/transcriptome/genes.gtf'
CDSs = gtf.get_CDSs(gtf_fn)
gtf_dict = {t.name: t for t in CDSs}
genes = read_nagalakshmi_file('nagalakshmi_annotations.txt')
discrepancies_after = {}
for name in genes:
if name not in gtf_dict:
print name, 'not in gtf_dict'
continue
start, end = genes[name]['SGD_Start'] - 1, genes[name]['SGD_End'] - 1
chrom = genes[name]['Chrom']
if start > end:
start, end = end, start
if start != gtf_dict[name].start or end != gtf_dict[name].end:
#print name, chrom
#print start, end
#print gtf_dict[name].start, gtf_dict[name].end
#print
if chrom not in discrepancies_after:
discrepancies_after[chrom] = start
else:
discrepancies_after[chrom] = min(start,
discrepancies_after[chrom])
return discrepancies_after
def sanity_check_nagalakshmi_file():
gtf_fn = '/home/jah/projects/arlen/data/organisms/saccharomyces_cerevisiae/SGD1.01/transcriptome/genes.gtf'
CDSs = gtf.get_CDSs(gtf_fn)
gtf_dict = {t.name: t for t in CDSs}
genes = read_nagalakshmi_file('nagalakshmi_annotations.txt')
discrepancies_after = {}
for name in genes:
if name not in gtf_dict:
print name, 'not in gtf_dict'
continue
start, end = genes[name]['SGD_Start'] - 1, genes[name]['SGD_End'] - 1
chrom = genes[name]['Chrom']
if start > end:
start, end = end, start
if start != gtf_dict[name].start or end != gtf_dict[name].end:
#print name, chrom
#print start, end
#print gtf_dict[name].start, gtf_dict[name].end
#print
if chrom not in discrepancies_after:
discrepancies_after[chrom] = start
else:
discrepancies_after[chrom] = min(start, discrepancies_after[chrom])
return discrepancies_after
def sanity_check_weinberg_file():
gtf_fn = '/home/jah/projects/arlen/data/organisms/saccharomyces_cerevisiae/EF4/transcriptome/genes.gtf'
CDSs = gtf.get_CDSs(gtf_fn)
gtf_dict = {t.name: t for t in CDSs}
genes = read_weinberg_file()
discrepancies_after = defaultdict(list)
for name in genes:
if name not in gtf_dict:
print name, 'not in gtf_dict'
continue
start, end = genes[name]['CdsStart'], genes[name]['CdsEnd']
chrom = genes[name]['Chromosome']
end -= 1
if start != gtf_dict[name].start or end != gtf_dict[name].end:
print name, chrom
print start, end
print gtf_dict[name].start, gtf_dict[name].end
raw_input()
discrepancies_after[chrom].append(name)
return discrepancies_after
def sanity_check_weinberg_file():
gtf_fn = '/home/jah/projects/arlen/data/organisms/saccharomyces_cerevisiae/EF4/transcriptome/genes.gtf'
CDSs = gtf.get_CDSs(gtf_fn)
gtf_dict = {t.name: t for t in CDSs}
genes = read_weinberg_file()
discrepancies_after = defaultdict(list)
for name in genes:
if name not in gtf_dict:
print name, 'not in gtf_dict'
continue
start, end = genes[name]['CdsStart'], genes[name]['CdsEnd']
chrom = genes[name]['Chromosome']
end -= 1
if start != gtf_dict[name].start or end != gtf_dict[name].end:
print name, chrom
print start, end
print gtf_dict[name].start, gtf_dict[name].end
raw_input()
discrepancies_after[chrom].append(name)
return discrepancies_after
def plot_frameshifts(
gtf_fn,
bam_fns,
gene_name,
exp_name,
genome_dir,
show_fractions=False,
):
codon_buffer = 10
start_codon = 'ATG'
stop_codons = {'TAA', 'TAG', 'TGA'}
A_site_offset = 5
CDSs = {c.name: c for c in gtf.get_CDSs(gtf_fn, genome_dir, '/dev/null')}
lengths = [28]
transcript = CDSs[gene_name]
left_buffer = 30 + codon_buffer * 3
right_buffer = (codon_buffer + 1) * 3
transcript.build_extent_maps(left_buffer, right_buffer)
experiment_counts = []
for bam_fn in bam_fns:
counts = positions.get_Transcript_extent_position_counts(
transcript,
bam_fn,
lengths,
left_buffer=left_buffer,
right_buffer=right_buffer,
)
experiment_counts.append(counts)
# Get the sequence of the extent.
extent_sequence = transcript.get_extent_sequence(
left_buffer=left_buffer,
right_buffer=right_buffer,
)
for length in lengths:
A_site_offset = positions.A_site_offsets['yeast'][length]
length_counts = reduce(
operator.add, [counts[length] for counts in experiment_counts])
codon_numbers = np.arange(-codon_buffer,
transcript.extent_length / 3 + codon_buffer)
# frame_counts_list[i, j] will be the number of RPF's starting at frame i of
# codon j
frame_counts_list = np.zeros((3, len(codon_numbers)), int)
start_codon_locations = [[] for _ in range(3)]
stop_codon_locations = [[] for _ in range(3)]
for c, codon_number in enumerate(codon_numbers):
codon_start = 3 * codon_number
for frame in range(3):
frame_counts_list[frame,
c] = length_counts['start', codon_start +
frame - A_site_offset]
codon = extent_sequence['start', codon_start +
frame:codon_start + frame + 3]
codon = ''.join(codon)
if codon == start_codon:
start_codon_locations[frame].append(codon_number)
if codon in stop_codons:
stop_codon_locations[frame].append(codon_number)
if show_fractions:
fig, axs = plt.subplots(4, 1, sharex=True)
cumulative_ax = axs[0]
frame_axs = axs[1:]
else:
fig, frame_axs = plt.subplots(3, 1, sharex=True)
for frame, (ax, frame_counts) in enumerate(
zip(frame_axs, frame_counts_list)):
nonzero_codon_numbers = [
c_n for c_n, f_c in zip(codon_numbers, frame_counts)
if f_c != 0
]
nonzero_frame_counts = [
f_c for c_n, f_c in zip(codon_numbers, frame_counts)
if f_c != 0
]
ax.plot(nonzero_codon_numbers, nonzero_frame_counts, '.')
ax.set_ylim(0, frame_counts_list.max() + 1)
ax.set_xlim(codon_numbers[0], codon_numbers[-1])
ax.set_title('Frame {0}'.format(frame))
ax.set_ylabel('Read counts')
for x in start_codon_locations[frame]:
ax.axvspan(x - 0.5,
x + 0.5,
facecolor='green',
edgecolor='none',
alpha=0.2)
for x in stop_codon_locations[frame]:
ax.axvspan(x - 0.5,
x + 0.5,
facecolor='red',
edgecolor='none',
alpha=0.2)
frame_axs[-1].set_xlabel('Codons from start codon')
if show_fractions:
frames_so_far = frame_counts_list.cumsum(axis=1)
fraction_frames_so_far = np.true_divide(
frames_so_far,
np.maximum(1, frames_so_far.sum(axis=0)),
)
frames_remaining = np.fliplr(
np.fliplr(frame_counts_list).cumsum(axis=1))
fraction_frames_remaining = np.true_divide(
frames_remaining,
np.maximum(1, frames_remaining.sum(axis=0)),
)
for frame in [0, 1, 2]:
so_far = fraction_frames_so_far[frame]
remaining = fraction_frames_remaining[frame]
color = colors[frame]
cumulative_ax.plot(codon_numbers,
so_far,
color=color,
label='{0} so far'.format(frame))
cumulative_ax.plot(codon_numbers,
remaining,
color=color,
linestyle='--',
label='{0} remaining'.format(frame))
cumulative_ax.set_xlim(codon_numbers[0])
cumulative_ax.set_ylim(-0.02, 1.02)
cumulative_ax.set_ylabel('Fraction of reads in extent')
cumulative_ax.legend(loc='upper right', framealpha=0.5)
fig.suptitle('{2}\n{0}\nlength {1} fragments'.format(
gene_name, length, exp_name))
return fig
def plot_frameshifts(gtf_fn, bam_fns, gene_name, exp_name, genome_dir, show_fractions=False):
codon_buffer = 10
start_codon = "ATG"
stop_codons = {"TAA", "TAG", "TGA"}
A_site_offset = 5
CDSs = {c.name: c for c in gtf.get_CDSs(gtf_fn, genome_dir, "/dev/null")}
lengths = [28]
transcript = CDSs[gene_name]
left_buffer = 30 + codon_buffer * 3
right_buffer = (codon_buffer + 1) * 3
transcript.build_extent_maps(left_buffer, right_buffer)
experiment_counts = []
for bam_fn in bam_fns:
counts = positions.get_Transcript_extent_position_counts(
transcript, bam_fn, lengths, left_buffer=left_buffer, right_buffer=right_buffer
)
experiment_counts.append(counts)
# Get the sequence of the extent.
extent_sequence = transcript.get_extent_sequence(left_buffer=left_buffer, right_buffer=right_buffer)
for length in lengths:
A_site_offset = positions.A_site_offsets["yeast"][length]
length_counts = reduce(operator.add, [counts[length] for counts in experiment_counts])
codon_numbers = np.arange(-codon_buffer, transcript.extent_length / 3 + codon_buffer)
# frame_counts_list[i, j] will be the number of RPF's starting at frame i of
# codon j
frame_counts_list = np.zeros((3, len(codon_numbers)), int)
start_codon_locations = [[] for _ in range(3)]
stop_codon_locations = [[] for _ in range(3)]
for c, codon_number in enumerate(codon_numbers):
codon_start = 3 * codon_number
for frame in range(3):
frame_counts_list[frame, c] = length_counts["start", codon_start + frame - A_site_offset]
codon = extent_sequence["start", codon_start + frame : codon_start + frame + 3]
codon = "".join(codon)
if codon == start_codon:
start_codon_locations[frame].append(codon_number)
if codon in stop_codons:
stop_codon_locations[frame].append(codon_number)
if show_fractions:
fig, axs = plt.subplots(4, 1, sharex=True)
cumulative_ax = axs[0]
frame_axs = axs[1:]
else:
fig, frame_axs = plt.subplots(3, 1, sharex=True)
for frame, (ax, frame_counts) in enumerate(zip(frame_axs, frame_counts_list)):
nonzero_codon_numbers = [c_n for c_n, f_c in zip(codon_numbers, frame_counts) if f_c != 0]
nonzero_frame_counts = [f_c for c_n, f_c in zip(codon_numbers, frame_counts) if f_c != 0]
ax.plot(nonzero_codon_numbers, nonzero_frame_counts, ".")
ax.set_ylim(0, frame_counts_list.max() + 1)
ax.set_xlim(codon_numbers[0], codon_numbers[-1])
ax.set_title("Frame {0}".format(frame))
ax.set_ylabel("Read counts")
for x in start_codon_locations[frame]:
ax.axvspan(x - 0.5, x + 0.5, facecolor="green", edgecolor="none", alpha=0.2)
for x in stop_codon_locations[frame]:
ax.axvspan(x - 0.5, x + 0.5, facecolor="red", edgecolor="none", alpha=0.2)
frame_axs[-1].set_xlabel("Codons from start codon")
if show_fractions:
frames_so_far = frame_counts_list.cumsum(axis=1)
fraction_frames_so_far = np.true_divide(frames_so_far, np.maximum(1, frames_so_far.sum(axis=0)))
frames_remaining = np.fliplr(np.fliplr(frame_counts_list).cumsum(axis=1))
fraction_frames_remaining = np.true_divide(frames_remaining, np.maximum(1, frames_remaining.sum(axis=0)))
for frame in [0, 1, 2]:
so_far = fraction_frames_so_far[frame]
remaining = fraction_frames_remaining[frame]
color = colors[frame]
cumulative_ax.plot(codon_numbers, so_far, color=color, label="{0} so far".format(frame))
cumulative_ax.plot(
codon_numbers, remaining, color=color, linestyle="--", label="{0} remaining".format(frame)
)
cumulative_ax.set_xlim(codon_numbers[0])
cumulative_ax.set_ylim(-0.02, 1.02)
cumulative_ax.set_ylabel("Fraction of reads in extent")
cumulative_ax.legend(loc="upper right", framealpha=0.5)
fig.suptitle("{2}\n{0}\nlength {1} fragments".format(gene_name, length, exp_name))
return fig
