def spectra_grid(json_path, group_label, plot_cfg, no_type3, figpath, format,
sample_size, force_overwrite, dry_run):
"""draws logo from mutation spectra analysis"""
# the following is for logging
args = locals()
if no_type3:
util.exclude_type3_fonts()
if not figpath:
dirname = os.path.dirname(json_path)
figpath = os.path.join(dirname, "spectra_grid.%s" % format)
log_file_path = os.path.join(dirname, "spectra_grid.log")
else:
figpath = util.abspath(figpath)
log_file_path = "%s.log" % ".".join(figpath.split(".")[:-1])
LOGGER.log_file_path = log_file_path
LOGGER.log_message(str(args), label='vars')
data = load_spectra_data(json_path, group_label)
if plot_cfg:
LOGGER.input_file(plot_cfg)
plot_cfg = util.get_plot_configs(cfg_path=plot_cfg)
f = draw_spectrum_grid(data, sample_size=sample_size, plot_cfg=plot_cfg)
f.savefig(figpath)
LOGGER.output_file(figpath)
click.secho("Wrote %s" % figpath, fg="green")
def spectra(countsfile, outpath, countsfile2, strand_symmetry, force_overwrite,
no_type3, dry_run, verbose):
'''log-linear analysis of mutation spectra between groups
'''
if no_type3:
util.exclude_type3_fonts()
spectra_analysis.main(countsfile, outpath,
countsfile2, strand_symmetry,
force_overwrite, dry_run,
verbose)
def nbr_matrix(paths_cfg, plot_cfg, figpath, format, no_type3, sample_size,
force_overwrite, dry_run):
'''draws square matrix of sequence logo's from neighbour analysis'''
if no_type3:
util.exclude_type3_fonts()
args = locals()
LOGGER.log_message(str(args), label='vars')
config_path = util.abspath(paths_cfg)
indir = os.path.dirname(config_path)
parser = SafeConfigParser()
parser.optionxform = str # stops automatic conversion to lower case
parser.read(config_path)
json_paths = {}
for direction, path in parser.items("json_paths"):
# assumes paths are relative to indir
path = os.path.join(indir, path)
if not os.path.exists(path):
print("Couldn't find %s" % path)
print("json file paths should be relative to paths_cfg")
sys.exit(1)
json_paths[direction] = path
if not figpath:
figpath = os.path.join(indir, "nbr_matrix.%s" % format)
log_file_path = os.path.join(indir, "nbr_matrix.log")
else:
figpath = util.abspath(figpath)
log_file_path = "%s.log" % ".".join(figpath.split(".")[:-1])
LOGGER.log_file_path = log_file_path
plot_data = {}
for direction, path in list(json_paths.items()):
LOGGER.input_file(path)
data = util.load_loglin_stats(path)
plot_data[direction] = data
fig = draw_position_grid(plot_data, sample_size)
fig.text(0.4, 0.955, "Ending Base", fontsize=20)
fig.text(0.03, 0.55, "Starting Base", rotation=90, fontsize=20)
fig.tight_layout(rect=(0.06, 0, 0.95, 0.95))
fig.savefig(figpath)
LOGGER.output_file(figpath)
click.secho("Wrote %s" % figpath, fg="green")
def mi(json_path, plot_cfg, no_type3, figpath, format, sample_size,
force_overwrite, dry_run):
"""draws conventional sequence logo, using MI, from first order effects"""
# the following is for logging
json_path = util.abspath(json_path)
args = locals()
if no_type3:
util.exclude_type3_fonts()
if not figpath:
dirname = os.path.dirname(json_path)
figpath = os.path.join(dirname, "MI.%s" % format)
log_file_path = os.path.join(dirname, "MI.log")
else:
figpath = util.abspath(figpath)
log_file_path = "%s.log" % ".".join(figpath.split(".")[:-1])
LOGGER.log_file_path = log_file_path
if plot_cfg:
LOGGER.input_file(plot_cfg)
LOGGER.log_message(str(args), label='vars')
data = util.load_loglin_stats(json_path)
positions = list(data.keys())
positions.sort()
num_pos = len(positions) + 1
mp = num_pos // 2
counts_array = numpy.zeros((4, num_pos), int)
for i, pos in enumerate(positions):
if i >= mp:
i += 1
pos_stats = data[pos]['stats']
counts = pos_stats[pos_stats['mut'] == 'M'][["base", "count"]]
counts = dict(zip(counts['base'], counts['count']))
for base in counts:
base_index = DNA.alphabet.index(base)
counts_array[base_index, i] = counts[base]
freq_matrix = entropy.counts_to_freq_matrix(counts_array)
mit = entropy.get_mit(freq_matrix, freq_matrix=True)
mi = mit.sum(axis=0)
char_hts = get_mi_char_heights(numpy.fabs(mit), mi)
plot_cfg = util.get_plot_configs(cfg_path=plot_cfg)
figsize = plot_cfg.get('1-way plot', 'figsize')
ytick_font = plot_cfg.get('1-way plot', 'ytick_fontsize')
xtick_font = plot_cfg.get('1-way plot', 'xtick_fontsize')
ylabel_font = plot_cfg.get('1-way plot', 'ylabel_fontsize')
xlabel_font = plot_cfg.get('1-way plot', 'xlabel_fontsize')
fig = logo.draw_multi_position(char_hts.T,
characters=[list(DNA)] * num_pos,
position_indices=list(range(num_pos)),
figsize=figsize,
figwidth=figsize[0],
xtick_fontsize=xtick_font,
ytick_fontsize=ytick_font,
sort_data=True)
ax = fig.gca()
ax.tick_params(axis='y', labelsize=ytick_font)
ax.tick_params(axis='x', labelsize=xtick_font)
ax.set_ylabel("MI", fontsize=ylabel_font)
ax.set_xlabel("Position", fontsize=xlabel_font)
fig.tight_layout()
fig.savefig(figpath)
LOGGER.output_file(figpath)
click.secho("Wrote %s" % figpath, fg="green")
def grid(fig_config, figpath, format, no_type3):
"""draws an arbitrary shaped grid of mutation motifs based on fig_config"""
# we read in the config file and determine number of rows and columns
# paths, headings, etc ..
# then create the figure and axes and call the mutation_motif drawing code
args = locals()
if no_type3:
util.exclude_type3_fonts()
if not figpath:
dirname = os.path.dirname(fig_config.name)
figpath = os.path.join(dirname, "drawn_array.%s" % format)
log_file_path = os.path.join(dirname, "drawn_array.log")
else:
figpath = util.abspath(figpath)
log_file_path = "%s.log" % ".".join(figpath.split(".")[:-1])
util.makedirs(os.path.dirname(figpath))
LOGGER.log_file_path = log_file_path
LOGGER.log_message(str(args), label='vars')
ncols, nrows, figsize, col_labels, row_labels, paths, axis_cfg = \
read_plot_array_config(fig_config)
print("ncols:", ncols)
print("nrows:", nrows)
print("figsize:", figsize)
print("col_labels:", col_labels)
print("row_labels:", row_labels)
print("paths:", paths)
print("axis_cfg:", axis_cfg)
#TODO: Convert below into Cogent3 Plotly
#-Plotly
layout = UnionDict(shapes=[])
adaptive_y = 0
plottable = {}
for coord in paths:
data = util.load_loglin_stats(paths[coord])
positions = list(data)
positions.sort()
heights, characters, indices = get_plot_data(data, positions)
adaptive_y = max(adaptive_y, logo.est_ylim(heights))
plottable[coord] = dict(char_heights=heights,
characters=characters,
position_indices=indices)
ylim = axis_cfg.get("ylim", adaptive_y)
for coord in plottable:
kwargs = plottable[coord]
kwargs["ax"] = coord
kwargs["ylim"] = ylim
r = logo.draw_multi_position_cogent3(**kwargs)
for key in r:
if key == "shapes":
layout.shapes.extend(r.shapes)
else:
layout[key] = r[key]
for i in range(0, ncols):
xaxis = "xaxis" + str(i + 1 if i != 0 else "")
layout[xaxis]["domain"] = [
0.0 + (i * (1 / ncols)), (i * (1 / ncols)) + (1 / ncols)
]
print(layout)
MARGININCHES = 0
PPI = 100
fig = Drawable(layout=layout,
width=(figsize[0] - MARGININCHES) * PPI,
height=(figsize[1] - MARGININCHES) * PPI)
#export
fig.write(path=figpath)
click.secho("Wrote Cogent3 %s" % figpath, fg="green")
"""
def grid(fig_config, figpath, format, no_type3):
"""draws an arbitrary shaped grid of mutation motifs based on fig_config"""
# we read in the config file and determine number of rows and columns
# paths, headings, etc ..
# then create the figure and axes and call the mutation_motif drawing code
args = locals()
if no_type3:
util.exclude_type3_fonts()
if not figpath:
dirname = os.path.dirname(fig_config.name)
figpath = os.path.join(dirname, "drawn_array.%s" % format)
log_file_path = os.path.join(dirname, "drawn_array.log")
else:
figpath = util.abspath(figpath)
log_file_path = "%s.log" % ".".join(figpath.split(".")[:-1])
util.makedirs(os.path.dirname(figpath))
LOGGER.log_file_path = log_file_path
LOGGER.log_message(str(args), label='vars')
ncols, nrows, figsize, col_labels, row_labels, paths, axis_cfg = \
read_plot_array_config(fig_config)
fig, axes = pyplot.subplots(nrows=nrows, ncols=ncols, figsize=figsize,
sharex=True, sharey=True)
figwidth = fig.get_figwidth()
try:
axes[0]
except TypeError:
axes = numpy.array([[axes]])
if len(axes.shape) == 1:
# required for indexing of appropriate axis
axes = numpy.vstack(axes)
if nrows == 1:
axes = axes.T
adaptive_y = 0
plottable = {}
for coord in paths:
data = util.load_loglin_stats(paths[coord])
positions = list(data)
positions.sort()
heights, characters, indices = get_plot_data(data, positions)
adaptive_y = max(adaptive_y, logo.est_ylim(heights))
plottable[coord] = dict(char_heights=heights,
characters=characters,
position_indices=indices,
figwidth=figwidth,
verbose=False)
ylim = axis_cfg.get("ylim", adaptive_y)
for coord in plottable:
kwargs = plottable[coord]
kwargs["ax"] = axes[coord]
kwargs["ylim"] = ylim
fig = logo.draw_multi_position(**kwargs)
xformat = FuncFormatter(format_float(1e-3, float_places=2))
for col in range(ncols):
top_ax = axes[0, col]
top_ax.set_title(col_labels[col], fontsize=axis_cfg["xlabel_fontsize"],
weight="bold", y=1.1)
btm_ax = axes[-1, col]
for xticklabel in btm_ax.get_xticklabels():
xticklabel.set_fontsize(axis_cfg["xtick_fontsize"])
xticklabel.set_rotation(0)
btm_ax.set_xlabel("Position", fontsize=axis_cfg["xlabel_fontsize"],
weight="bold")
btm_ax.xaxis.labelpad = axis_cfg['xlabel_pad']
for row in range(nrows):
lft_ax = axes[row, 0]
for yticklabel in lft_ax.get_yticklabels():
yticklabel.set_fontsize(axis_cfg["ytick_fontsize"])
yticklabel.set_rotation(0)
lft_ax.yaxis.set_major_formatter(FuncFormatter(xformat))
lft_ax.yaxis.labelpad = axis_cfg['ylabel_pad']
lft_ax.set_ylabel(row_labels[row], rotation=0,
fontsize=axis_cfg['ylabel_fontsize'],
weight="bold")
fig.tight_layout()
fig.savefig(figpath)
click.secho("Wrote %s" % figpath, fg="green")
def nbr(countsfile, outpath, countsfile2, first_order, strand_symmetry,
group_label, group_ref, plot_cfg, no_type3, format, verbose, dry_run):
'''log-linear analysis of neighbouring base influence on point mutation
Writes estimated statistics, figures and a run log to the specified
directory outpath.
See documentation for count table format requirements.
'''
if no_type3:
util.exclude_type3_fonts()
args = locals()
outpath = util.abspath(outpath)
if not dry_run:
util.makedirs(outpath)
runlog_path = os.path.join(outpath, "analysis.log")
LOGGER.log_file_path = runlog_path
LOGGER.log_message(str(args), label='vars')
counts_filename = util.abspath(countsfile)
counts_table = util.load_table_from_delimited_file(counts_filename,
sep='\t')
LOGGER.input_file(counts_filename, label="countsfile1_path")
positions = [c for c in counts_table.header if c.startswith('pos')]
if not first_order and len(positions) != 4:
raise ValueError("Requires four positions for analysis")
group_label = group_label or None
group_ref = group_ref or None
if strand_symmetry:
group_label = 'strand'
group_ref = group_ref or '+'
if group_label not in counts_table.header:
print("ERROR: no column named 'strand', exiting.")
exit(-1)
if countsfile2:
print("Performing 2 group analysis")
group_label = group_label or 'group'
group_ref = group_ref or '1'
counts_table1 = counts_table.with_new_column(group_label,
lambda x: '1',
columns=counts_table.header[0])
fn2 = util.abspath(countsfile2)
counts_table2 = util.load_table_from_delimited_file(fn2, sep='\t')
LOGGER.input_file(fn2, label="countsfile2_path")
counts_table2 = counts_table2.with_new_column(group_label,
lambda x: '2',
columns=counts_table2.header[0])
# now combine
header = [group_label] + counts_table2.header[:-1]
raw1 = counts_table1.tolist(header)
raw2 = counts_table2.tolist(header)
counts_table = make_table(header=header, rows=raw1 + raw2)
if not dry_run:
outfile = os.path.join(outpath, 'group_counts_table.txt')
counts_table.write(outfile, sep='\t')
LOGGER.output_file(outfile, label="group_counts")
if dry_run or verbose:
print()
print(counts_table)
print()
plot_config = util.get_plot_configs(cfg_path=plot_cfg)
msg = single_group(counts_table, outpath, group_label, group_ref,
positions, plot_config, first_order,
dry_run)
print(msg)