comparison cwpair2_util.py @ 8:1fc26b8e618d draft

Uploaded

7:d455f14530dc

import traceback
import matplotlib
matplotlib.use('Agg')
from matplotlib import pyplot

DETAILS = 'D'
FINAL_PLOTS = 'F'
ORPHANS = 'O'
PREVIEW_PLOTS = 'P'
SIMPLES = 'S'
STATS_GRAPH = 'C'
GFF_EXT = 'gff'
TABULAR_EXT = 'tabular'

# Graph settings.
COLORS = 'krg'
Y_LABEL = 'Peak-pair counts'
X_LABEL = 'Peak-pair distance (bp)'

    for l in ax.get_xticklines() + ax.get_yticklines():
        l.set_markeredgewidth(TICK_WIDTH)
    pyplot.savefig(fname)


def create_directories(method):
    if method == 'all':
        match_methods = METHODS.keys()
    else:
        match_methods = [method]
    for match_method in match_methods:
        os.mkdir('%s_%s' % (match_method, DETAILS))
        os.mkdir('%s_%s' % (match_method, FINAL_PLOTS))
        os.mkdir('%s_%s' % (match_method, ORPHANS))
        os.mkdir('%s_%s' % (match_method, PREVIEW_PLOTS))
        os.mkdir('%s_%s' % (match_method, SIMPLES))
        os.mkdir('%s_%s' % (match_method, STATS_GRAPH))


def process_file(dataset_path, galaxy_hid, method, threshold, up_distance,
                 down_distance, binsize, output_files, sort_score):
    if method == 'all':
        match_methods = METHODS.keys()
    else:
        match_methods = [method]
    statistics = []

                                match_method,
                                threshold,
                                up_distance,
                                down_distance,
                                binsize,
                                output_files,
                                sort_score)
        statistics.append(stats)
    if output_files == 'all' and method == 'all':
        frequency_plot([s['dist'] for s in statistics],
                       statistics[0]['graph_path'],
                       labels=METHODS.keys())
    return statistics


def perform_process(dataset_path, galaxy_hid, method, threshold, up_distance,
                    down_distance, binsize, output_files, sort_score):
    output_details = output_files in ["all", "simple_orphan_detail"]
    output_plots = output_files in ["all"]
    output_orphans = output_files in ["all", "simple_orphan", "simple_orphan_detail"]
    # Keep track of statistics for the output file
    statistics = {}
    input = csv.reader(open(dataset_path, 'rt'), delimiter='\t')
    fpath, fname = os.path.split(dataset_path)
    statistics['fname'] = '%s: data %s' % (method, str(galaxy_hid))
    statistics['dir'] = fpath
    if threshold >= 1:
        filter_string = 'fa%d' % threshold
    else:
        filter_string = 'f%d' % (threshold * 100)
    fname = 'data_%s_%su%dd%db%d' % (galaxy_hid, filter_string, up_distance, down_distance, binsize)

    def make_path(output_type, extension=TABULAR_EXT):
        # Returns the full path for a certain output.
        return os.path.join(output_type, '%s_%s.%s' % (output_type, fname, extension))

    def td_writer(output_type, extension=TABULAR_EXT):
        # Returns a tab-delimited writer for a specified output.
        output_file_path = make_path(output_type, extension)
        return csv.writer(open(output_file_path, 'wt'), delimiter='\t')

    try:
        chromosomes = parse_chromosomes(input)
    except Exception:
        stop_err('Unable to parse file "%s".\n%s' % (dataset_path, traceback.format_exc()))
    if output_details:
        # Details
        detailed_output = td_writer('%s_%s' % (method, DETAILS), extension=TABULAR_EXT)
        detailed_output.writerow(('chrom', 'start', 'end', 'value', 'strand') * 2 + ('midpoint', 'c-w reads sum', 'c-w distance (bp)'))
    if output_plots:
        # Final Plot
        final_plot_path = make_path('%s_%s' % (method, FINAL_PLOTS), PLOT_FORMAT)
    if output_orphans:
        # Orphans
        orphan_output = td_writer('%s_%s' % (method, ORPHANS), extension=TABULAR_EXT)
        orphan_output.writerow(('chrom', 'strand', 'start', 'end', 'value'))
    if output_plots:
        # Preview Plot
        preview_plot_path = make_path('%s_%s' % (method, PREVIEW_PLOTS), PLOT_FORMAT)
    # Simple
    simple_output = td_writer('%s_%s' % (method, SIMPLES), extension=GFF_EXT)
    statistics['stats_path'] = 'statistics.%s' % TABULAR_EXT
    if output_plots:
        statistics['graph_path'] = make_path('%s_%s' % (method, STATS_GRAPH), PLOT_FORMAT)
    statistics['perc95'] = perc95(chromosomes)
    if threshold > 0:
        # Apply filter
        filter(chromosomes, threshold)
    if method == 'mode':

        if output_orphans:
            for cpeak in crick:
                orphan_output.writerow((cname, cpeak[0], cpeak[1], cpeak[2], cpeak[3]))
        # Keep track of orphans for statistics.
        orphans += len(crick)
    # Sort output by score if specified.
    if sort_score == "desc":
        x.sort(key=lambda data: float(data[5]), reverse=True)
    elif sort_score == "asc":
        x.sort(key=lambda data: float(data[5]))
    # Writing a summary to gff format file
    for row in x:
        row_tmp = list(row)
        # Dataset in tuple cannot be modified in Python, so row will
        # be converted to list format to add 'chr'.

        elif row_tmp[0] == "997":
            row_tmp[0] = 'chrX'
        else:
            row_tmp[0] = row_tmp[0]
        # Print row_tmp.
        simple_output.writerow(row_tmp)
    statistics['paired'] = dist.size() * 2
    statistics['orphans'] = orphans
    statistics['final_mode'] = dist.mode()
    if output_plots:
        frequency_plot([dist], final_plot_path, title='Frequency distribution')

8:1fc26b8e618d

import traceback
import matplotlib
matplotlib.use('Agg')
from matplotlib import pyplot

# Data outputs
DETAILS = 'D'
MATCHED_PAIRS = 'MP'
ORPHANS = 'O'
# Data output formats
GFF_EXT = 'gff'
TABULAR_EXT = 'tabular'
# Statistics historgrams output directory.
HISTOGRAM = 'H'
# Statistics outputs
FINAL_PLOTS = 'F'
PREVIEW_PLOTS = 'P'
STATS_GRAPH = 'C'

# Graph settings.
COLORS = 'krg'
Y_LABEL = 'Peak-pair counts'
X_LABEL = 'Peak-pair distance (bp)'

    for l in ax.get_xticklines() + ax.get_yticklines():
        l.set_markeredgewidth(TICK_WIDTH)
    pyplot.savefig(fname)


def create_directories():
    # Output histograms in pdf.
    os.mkdir(HISTOGRAM)
    os.mkdir('data_%s' % DETAILS)
    os.mkdir('data_%s' % ORPHANS)
    os.mkdir('data_%s' % MATCHED_PAIRS)


def process_file(dataset_path, galaxy_hid, method, threshold, up_distance,
                 down_distance, binsize, output_files):
    if method == 'all':
        match_methods = METHODS.keys()
    else:
        match_methods = [method]
    statistics = []

                                match_method,
                                threshold,
                                up_distance,
                                down_distance,
                                binsize,
                                output_files)
        statistics.append(stats)
    if output_files == 'all' and method == 'all':
        frequency_plot([s['dist'] for s in statistics],
                       statistics[0]['graph_path'],
                       labels=METHODS.keys())
    return statistics


def perform_process(dataset_path, galaxy_hid, method, threshold, up_distance,
                    down_distance, binsize, output_files):
    output_details = output_files in ["all", "matched_pair_orphan_detail"]
    output_plots = output_files in ["all"]
    output_orphans = output_files in ["all", "matched_pair_orphan", "matched_pair_orphan_detail"]
    # Keep track of statistics for the output file
    statistics = {}
    input = csv.reader(open(dataset_path, 'rt'), delimiter='\t')
    fpath, fname = os.path.split(dataset_path)
    statistics['fname'] = '%s: data %s' % (method, str(galaxy_hid))
    statistics['dir'] = fpath
    if threshold >= 1:
        filter_string = 'fa%d' % threshold
    else:
        filter_string = 'f%d' % (threshold * 100)
    fname = '%s_%su%dd%d_on_data_%s' % (method, filter_string, up_distance, down_distance, galaxy_hid)

    def make_histogram_path(output_type, fname):
        return os.path.join(HISTOGRAM, 'histogram_%s_%s.%s' % (output_type, fname, PLOT_FORMAT))

    def make_path(output_type, extension, fname):
        # Returns the full path for an output.
        return os.path.join(output_type, '%s_%s.%s' % (output_type, fname, extension))

    def td_writer(output_type, extension, fname):
        # Returns a tab-delimited writer for a specified output.
        output_file_path = make_path(output_type, extension, fname)
        return csv.writer(open(output_file_path, 'wt'), delimiter='\t')

    try:
        chromosomes = parse_chromosomes(input)
    except Exception:
        stop_err('Unable to parse file "%s".\n%s' % (dataset_path, traceback.format_exc()))
    if output_details:
        # Details
        detailed_output = td_writer('data_%s' % DETAILS, TABULAR_EXT, fname)
        detailed_output.writerow(('chrom', 'start', 'end', 'value', 'strand') * 2 + ('midpoint', 'c-w reads sum', 'c-w distance (bp)'))
    if output_plots:
        # Final Plot
        final_plot_path = make_histogram_path(FINAL_PLOTS, fname)
    if output_orphans:
        # Orphans
        orphan_output = td_writer('data_%s' % ORPHANS, TABULAR_EXT, fname)
        orphan_output.writerow(('chrom', 'strand', 'start', 'end', 'value'))
    if output_plots:
        # Preview Plot
        preview_plot_path = make_histogram_path(PREVIEW_PLOTS, fname)
    # Matched Pairs.
    matched_pairs_output = td_writer('data_%s' % MATCHED_PAIRS, GFF_EXT, fname)
    statistics['stats_path'] = 'statistics.%s' % TABULAR_EXT
    if output_plots:
        statistics['graph_path'] = make_histogram_path(STATS_GRAPH, fname)
    statistics['perc95'] = perc95(chromosomes)
    if threshold > 0:
        # Apply filter
        filter(chromosomes, threshold)
    if method == 'mode':

        if output_orphans:
            for cpeak in crick:
                orphan_output.writerow((cname, cpeak[0], cpeak[1], cpeak[2], cpeak[3]))
        # Keep track of orphans for statistics.
        orphans += len(crick)
    # Sort output descending by score.
    x.sort(key=lambda data: float(data[5]), reverse=True)
    # Writing a summary to gff format file
    for row in x:
        row_tmp = list(row)
        # Dataset in tuple cannot be modified in Python, so row will
        # be converted to list format to add 'chr'.

        elif row_tmp[0] == "997":
            row_tmp[0] = 'chrX'
        else:
            row_tmp[0] = row_tmp[0]
        # Print row_tmp.
        matched_pairs_output.writerow(row_tmp)
    statistics['paired'] = dist.size() * 2
    statistics['orphans'] = orphans
    statistics['final_mode'] = dist.mode()
    if output_plots:
        frequency_plot([dist], final_plot_path, title='Frequency distribution')