GEO Series matrix to project and sample annotations

series_matrix2csv.py

def series_matrix2csv(matrix_url, prefix=None):
    """
    matrix_url: gziped URL with GEO series matrix.
    """
    import gzip
    import pandas as pd

    os.system("wget {}".format(matrix_url))
    filename = matrix_url.split("/")[-1]

    with gzip.open(filename, 'rb') as f:
        file_content = f.read()

    # separate lines with only one field (project-related)
    # from lines with >2 fields (sample-related)
    prj_lines = dict()
    sample_lines = dict()

    for line in file_content.decode("utf-8").strip().split("\n"):
        line = line.strip().split("\t")
        if len(line) == 2:
            prj_lines[line[0].replace("\"", "")] = line[1].replace("\"", "")
        elif len(line) > 2:
            sample_lines[line[0].replace("\"", "")] = [x.replace("\"", "") for x in line[1:]]

    prj = pd.Series(prj_lines)
    prj.index = prj.index.str.replace("!Series_", "")

    samples = pd.DataFrame(sample_lines)
    samples.columns = samples.columns.str.replace("!Sample_", "")

    if prefix is not None:
        prj.to_csv(os.path.join(prefix + ".project_annotation.csv"), index=True)
        samples.to_csv(os.path.join(prefix + ".sample_annotation.csv"), index=False)

    return prj, samples