comparison repmatch_gff3_util.py @ 8:d10ae3aeebc8 draft

Uploaded

7:1807688a8a5f

    os.close(fd)
    return name


def process_files(dataset_paths, galaxy_hids, method, distance, step, replicates, up_limit, low_limit, output_files,
                  output_summary, output_orphan, output_detail, output_key, output_histogram):
    output_histogram_file = output_files in ["all"] and method in ["all"]
    if len(dataset_paths) < 2:
        return
    if method == 'all':
        match_methods = METHODS.keys()
    else:

                                     step,
                                     replicates,
                                     up_limit,
                                     low_limit,
                                     output_files,
                                     output_summary,
                                     output_orphan,
                                     output_detail,
                                     output_key,
                                     output_histogram)
    if output_histogram_file:
        tmp_histogram_path = get_temporary_plot_path()
        frequency_histogram([stat['distribution'] for stat in [statistics]],
                            tmp_histogram_path,
                            METHODS.keys())
        shutil.move(tmp_histogram_path, output_histogram)


def perform_process(dataset_paths, galaxy_hids, method, distance, step, num_required, up_limit, low_limit, output_files,
                    output_summary, output_orphan, output_detail, output_key, output_histogram):
    output_detail_file = output_files in ["all"] and output_detail is not None
    output_key_file = output_files in ["all"] and output_key is not None
    output_orphan_file = output_files in ["all", "simple_orphan"] and output_orphan is not None
    output_histogram_file = output_files in ["all"] and output_histogram is not None
    replicates = []
    for i, dataset_path in enumerate(dataset_paths):
        try:
            galaxy_hid = galaxy_hids[i]
            r = Replicate(galaxy_hid, dataset_path)

                   'median midpoint',
                   'median midpoint+1',
                   'c-w sum',
                   'c-w distance',
                   'replicate id')
    summary_output = td_writer(output_summary)
    if output_key_file:
        key_output = td_writer(output_key)
        key_output.writerow(('data', 'median read count'))
    if output_detail_file:
        detail_output = td_writer(output_detail)
        detail_output.writerow(labels)
    if output_orphan_file:
        orphan_output = td_writer(output_orphan)
        orphan_output.writerow(('chrom', 'midpoint', 'midpoint+1', 'c-w sum', 'c-w distance', 'replicate id'))
    # Perform filtering
    if up_limit < 1000 or low_limit > -1000:
        for replicate in replicates:
            replicate.filter(up_limit, low_limit)
    # Actually merge the peaks
    peak_groups = []
    orphans = []
    freq = FrequencyDistribution()

    def do_match(reps, distance):
        # Copy list because we will mutate it, but keep replicate references.
        reps = reps[:]

    else:
        for d in range(0, distance, step):
            do_match(replicates, d)
    for group in peak_groups:
        freq.add(group.num_replicates)
    # Collect together the remaining orphans
    for replicate in replicates:
        for chromosome in replicate.chromosomes.values():
            for peak in chromosome.peaks:
                freq.add(1)
                orphans.append(peak)
    # Average the orphan count in the graph by # replicates
    med = median([peak.value for group in peak_groups for peak in group.peaks.values()])
    for replicate in replicates:
        replicate.median = median([peak.value for group in peak_groups for peak in group.peaks.values() if peak.replicate == replicate])
        key_output.writerow((replicate.id, replicate.median))
    for group in peak_groups:
        # Output summary (matched pairs).
        summary_output.writerow(gff_row(cname=group.chrom,
                                        start=group.midpoint,
                                        end=group.midpoint+1,
                                        source='repmatch',
                                        score=group.normalized_value(med),
                                        attrs={'median_distance': group.median_distance,
                                               'replicates': group.num_replicates,
                                               'value_sum': group.value_sum}))
        if output_detail_file:
            summary = (group.chrom,
                       group.midpoint,
                       group.midpoint+1,
                       group.normalized_value(med),
                       group.num_replicates,
                       group.median_distance,
                       group.value_sum)
            for peak in group.peaks.values():
                summary += (peak.chrom, peak.midpoint, peak.midpoint+1, peak.value, peak.distance, peak.replicate.id)
            detail_output.writerow(summary)
    if output_orphan_file:
        for orphan in orphans:
            orphan_output.writerow((orphan.chrom,
                                    orphan.midpoint,
                                    orphan.midpoint+1,
                                    orphan.value,
                                    orphan.distance,
                                    orphan.replicate.id))
    if output_histogram_file:
        tmp_histogram_path = get_temporary_plot_path()
        frequency_histogram([freq], tmp_histogram_path)
        shutil.move(tmp_histogram_path, output_histogram)
    return {'distribution': freq}

8:d10ae3aeebc8

    os.close(fd)
    return name


def process_files(dataset_paths, galaxy_hids, method, distance, step, replicates, up_limit, low_limit, output_files,
                  output_matched_peaks, output_unmatched_peaks, output_detail, output_statistics_table, output_statistics_histogram):
    output_statistics_histogram_file = output_files in ["all"] and method in ["all"]
    if len(dataset_paths) < 2:
        return
    if method == 'all':
        match_methods = METHODS.keys()
    else:

                                     step,
                                     replicates,
                                     up_limit,
                                     low_limit,
                                     output_files,
                                     output_matched_peaks,
                                     output_unmatched_peaks,
                                     output_detail,
                                     output_statistics_table,
                                     output_statistics_histogram)
    if output_statistics_histogram_file:
        tmp_statistics_histogram_path = get_temporary_plot_path()
        frequency_histogram([stat['distribution'] for stat in [statistics]],
                            tmp_statistics_histogram_path,
                            METHODS.keys())
        shutil.move(tmp_statistics_histogram_path, output_statistics_histogram)


def perform_process(dataset_paths, galaxy_hids, method, distance, step, num_required, up_limit, low_limit, output_files,
                    output_matched_peaks, output_unmatched_peaks, output_detail, output_statistics_table, output_statistics_histogram):
    output_detail_file = output_files in ["all"] and output_detail is not None
    output_statistics_table_file = output_files in ["all"] and output_statistics_table is not None
    output_unmatched_peaks_file = output_files in ["all", "matched_peaks_unmatched_peaks"] and output_unmatched_peaks is not None
    output_statistics_histogram_file = output_files in ["all"] and output_statistics_histogram is not None
    replicates = []
    for i, dataset_path in enumerate(dataset_paths):
        try:
            galaxy_hid = galaxy_hids[i]
            r = Replicate(galaxy_hid, dataset_path)

                   'median midpoint',
                   'median midpoint+1',
                   'c-w sum',
                   'c-w distance',
                   'replicate id')
    matched_peaks_output = td_writer(output_matched_peaks)
    if output_statistics_table_file:
        statistics_table_output = td_writer(output_statistics_table)
        statistics_table_output.writerow(('data', 'median read count'))
    if output_detail_file:
        detail_output = td_writer(output_detail)
        detail_output.writerow(labels)
    if output_unmatched_peaks_file:
        unmatched_peaks_output = td_writer(output_unmatched_peaks)
        unmatched_peaks_output.writerow(('chrom', 'midpoint', 'midpoint+1', 'c-w sum', 'c-w distance', 'replicate id'))
    # Perform filtering
    if up_limit < 1000 or low_limit > -1000:
        for replicate in replicates:
            replicate.filter(up_limit, low_limit)
    # Actually merge the peaks
    peak_groups = []
    unmatched_peaks = []
    freq = FrequencyDistribution()

    def do_match(reps, distance):
        # Copy list because we will mutate it, but keep replicate references.
        reps = reps[:]

    else:
        for d in range(0, distance, step):
            do_match(replicates, d)
    for group in peak_groups:
        freq.add(group.num_replicates)
    # Collect together the remaining unmatched_peaks
    for replicate in replicates:
        for chromosome in replicate.chromosomes.values():
            for peak in chromosome.peaks:
                freq.add(1)
                unmatched_peaks.append(peak)
    # Average the unmatched_peaks count in the graph by # replicates
    med = median([peak.value for group in peak_groups for peak in group.peaks.values()])
    for replicate in replicates:
        replicate.median = median([peak.value for group in peak_groups for peak in group.peaks.values() if peak.replicate == replicate])
        statistics_table_output.writerow((replicate.id, replicate.median))
    for group in peak_groups:
        # Output matched_peaks (matched pairs).
        matched_peaks_output.writerow(gff_row(cname=group.chrom,
                                              start=group.midpoint,
                                              end=group.midpoint+1,
                                              source='repmatch',
                                              score=group.normalized_value(med),
                                              attrs={'median_distance': group.median_distance,
                                                     'replicates': group.num_replicates,
                                                     'value_sum': group.value_sum}))
        if output_detail_file:
            matched_peaks = (group.chrom,
                             group.midpoint,
                             group.midpoint+1,
                             group.normalized_value(med),
                             group.num_replicates,
                             group.median_distance,
                             group.value_sum)
            for peak in group.peaks.values():
                matched_peaks += (peak.chrom, peak.midpoint, peak.midpoint+1, peak.value, peak.distance, peak.replicate.id)
            detail_output.writerow(matched_peaks)
    if output_unmatched_peaks_file:
        for unmatched_peak in unmatched_peaks:
            unmatched_peaks_output.writerow((unmatched_peak.chrom,
                                             unmatched_peak.midpoint,
                                             unmatched_peak.midpoint+1,
                                             unmatched_peak.value,
                                             unmatched_peak.distance,
                                             unmatched_peak.replicate.id))
    if output_statistics_histogram_file:
        tmp_statistics_histogram_path = get_temporary_plot_path()
        frequency_histogram([freq], tmp_statistics_histogram_path)
        shutil.move(tmp_statistics_histogram_path, output_statistics_histogram)
    return {'distribution': freq}