def download_idr_tfs(root_dir, metadata):
"""Download all tfs with idr called peaks"""
idr_records = fetch_idr_record(metadata)
## Theere is only one IDR per sample
if len(idr_records)!=1:
print(idr_records[0]['dataset'])
assert len(idr_records) <= 1
for idr_record in idr_records:
dataset = idr_record['dataset']
peakfilename = idr_record['peakfilename'] + '.bed.gz'
dataset_dir = os.path.join(root_dir, dataset)
safe_makedir(dataset_dir)
source_url = __base_url__ + idr_record['href']
print(source_url)
download_peakfile(source_url, peakfilename, dataset_dir)
save_metadata_json(idr_record, dataset_dir)
return {'assembly': idr_record['assembly'],'bedfile': os.path.join(dataset_dir, peakfilename.replace('.gz',''))}
def create_plot(meme_file,
plot_title,
output_dir=None,
centrimo_dir=None,
motif_number=1,
flank_length=5,
sample_score_files=[],
control_score_files=[],
reg_plot_titles=[],
annotate=None,
save=True):
"""Create plot
Parameters
----------
meme_file: string
Path to meme.txt
peak_file: string
Path to summit file
centrimo_dir: string
Path to centrimo's output directory
motif_number: int
1-based number of motif in the motif file
sample_score_files: list
Path to conservation scores files for sample
control_score_files: list
Path to conservation score files for control
legend_titles: list
List of legend titles
"""
meme_record = read_memefile(meme_file)
total_sequences = get_total_sequences(meme_file)
record = meme_record['motif_records'][motif_number-1]
num_occurrences = getattr(record, 'num_occurrences', 'Unknown')
all_meme_occurrences = []
for motif_record in meme_record['motif_records']:
all_meme_occurrences.append(getattr(motif_record, 'num_occurrences', 'Unknown'))
meme_dir = os.path.abspath(os.path.dirname(meme_file))
if not output_dir:
output_dir = os.path.join(os.path.join(meme_dir, '..'), 'moca_plots')
safe_makedir(output_dir)
subplot_ncols = 1
if len(sample_score_files) == 0:
raise MocaException('Found no sample score files')
elif len(control_score_files) == 0:
raise MocaException('Found no control score filees')
elif len(sample_score_files)!=len(control_score_files):
raise MocaException('Found unequal size of sample and control score files')
if annotate == "" or annotate == ' ':
annotate = None
subplot_ncols +=1
max_occur = get_max_occuring_bases(record, max_count=1, count_type=COUNT_TYPE)
motif_freq = []
for position in max_occur:
motif_freq.append(position[0][1])
motif_freq = np.asarray(motif_freq)
sample_conservation_scores = []
control_conservation_scores = []
for i in range(0, len(sample_score_files)):
sample_conservation_scores.append(np.loadtxt(sample_score_files[i]))
for i in range(0, len(control_score_files)):
control_conservation_scores.append(np.loadtxt(control_score_files[i]))
motif = record
motif_length = motif.length
motif_evalue = motif.evalue
meme_dir = os.path.abspath(os.path.dirname(meme_file))
X_values = [40+15] ## this is by trial and error, the position for the first base logo
## Generate all other X coordinates
for j in range(1,len(motif)+2*flank_length):
X_values.append( X_values[j-1]+OFFSET+1.9 )
if centrimo_dir:
subplot_ncols +=1
centrimo_dir = os.path.abspath(centrimo_dir)
centrimo_txt = os.path.join(centrimo_dir, 'centrimo.txt')
centrimo_stats = os.path.join(centrimo_dir, 'site_counts.txt')
plot_title += r' \# {}'.format(motif_number)
##FIXME This is a big dirty hacl to get thegenerate plots for the Reverse complement logo too
logo_name =['logo{}.png'.format(motif_number), 'logo_rc{}.png'.format(motif_number)]
figures = []
for sample_score, control_score, subplot_legend_title in zip(sample_conservation_scores,
control_conservation_scores,
reg_plot_titles):
for logo_filename in logo_name:
setup_matplotlib()
if 'rc'in logo_filename:
sample_score = sample_score[::-1]
matplot_dict = init_figure(meme_dir=meme_dir, X_values=X_values,
motif=motif_number,
subplot_ncols=subplot_ncols, annotate=annotate)
f = matplot_dict['figure']
gs = matplot_dict['gs']
figsize = matplot_dict['figsize']
right_margin = matplot_dict['right_margin']
#total_px= matplot_dict['total_px']
title = r'\textbf{' + '\\underline{'+'{}'.format(plot_title)+'}}'
f.suptitle(title, fontsize=LEGEND_FONTSIZE)
logo_plot = create_logo_plot({'figure':f, 'gridspec': gs[0]}, meme_dir, logo_filename, motif_length)
subgrid = gridspec.GridSpec(2, subplot_ncols, height_ratios=[1,2], width_ratios=[1]*subplot_ncols)
subgrid.update(bottom=0.14, right=0.9, left=1-right_margin*0.85, wspace=0.58)
X_left, X_center, X_right = create_stemplot({'figure': f,
'gridspec': gs[1],
'shareX': logo_plot},
X_values,
sample_score,
motif_length,
flank_length=flank_length,
legend_title=subplot_legend_title)
create_bar_plot(logo_plot, X_right, matplot_dict['height_px'],
total_sequences, all_meme_occurrences, motif_number, motif_evalue)
create_ols_legend_plot({'figure':f, 'gridspec': subgrid[0,0]}, motif_freq,
sample_score, control_score,
flank_length, legend_title=subplot_legend_title)
create_scatter_plot({'figure':f, 'gridspec': subgrid[1,0]}, motif_freq,
sample_score, control_score,
flank_length, num_occurrences, y_label=subplot_legend_title)
if centrimo_dir:
create_enrichment_plot({'figure': f,
'gridspec_header': subgrid[0,1],
'gridspec_body': subgrid[1,1]},
motif_number,
centrimo_txt,
centrimo_stats)
if 'rc' not in logo_filename:
out_file = os.path.join(output_dir,'moca_{}_{}.png'.format(subplot_legend_title, motif_number))
else:
out_file = os.path.join(output_dir,'moca_{}_{}_rc.png'.format(subplot_legend_title, motif_number))
if annotate:
create_annnotation_plot({'figure': f,
'gridspec_header': subgrid[0,-1],
'gridspec_body': subgrid[1,-1]},
annotate)
if save:
f.savefig(out_file, figsize=figsize, dpi=DPI)
figures.append(f)
plt.close('all')
return figures
