New upstream version 1.0.6

## 1.0.4 (in progress)

## 1.0.6

- Fix for the python3 fix.

## 1.0.5

- Fix for cb_filter with python3.

## 1.0.4

- Enable cb_histogram to be used on samples without UMIs.

- Enable filtering of cells during `demultiplex_cells`.

- Fix incorrect pandas.read_csv call with header=-1.

## 1.0.3 

- Python 3 support

setup(

        name='umis',

        version='1.0.3',

        version='1.0.6',

        description='Package for estimating UMI counts in Transcript Tag Counting data.',

        packages=find_packages(),

        install_requires=['click', 'pysam>=0.8.3', 'pandas', 'regex', 'scipy', 'toolz'],

import scipy.io, scipy.sparse

import click

VERSION = "1.0.3"

VERSION = "1.0.6"

logging.basicConfig(level=logging.INFO)

logger = logging.getLogger(__name__)

               "molecular": BarcodeInfo(bamtag="RX", readprefix="UMI")}

def open_gzipsafe(f):

    if is_python3():

        return gzip.open(f, mode="rt") if f.endswith(".gz") else open(f)

    else:

        return gzip.open(f) if f.endswith(".gz") else open(f)

def safe_makedir(dname):

    if filename == "-":

        filename_fh = sys.stdin

    elif filename.endswith('gz'):

        if is_python3:

        if is_python3():

            filename_fh = gzip.open(filename, mode='rt')

        else:

            filename_fh = BufferedReader(gzip.open(filename, mode='rt'))

    cb_hist = None

    filter_cb = False

    if cb_histogram:

        cb_hist = pd.read_csv(cb_histogram, index_col=0, header=-1, squeeze=True, sep="\t")

        cb_hist = pd.read_csv(cb_histogram, index_col=0, header=None, squeeze=True, sep="\t")

        total_num_cbs = cb_hist.shape[0]

        cb_hist = cb_hist[cb_hist > cb_cutoff]

        logger.info('Keeping {} out of {} cellular barcodes.'.format(cb_hist.shape[0], total_num_cbs))

    cb_hist = None

    filter_cb = False

    if cb_histogram:

        cb_hist = pd.read_csv(cb_histogram, index_col=0, header=-1, squeeze=True, sep="\t")

        cb_hist = pd.read_csv(cb_histogram, index_col=0, header=None, squeeze=True, sep="\t")

        total_num_cbs = cb_hist.shape[0]

        cb_hist = cb_hist[cb_hist > cb_cutoff]

        logger.info('Keeping {} out of {} cellular barcodes.'.format(cb_hist.shape[0], total_num_cbs))

    for read in read_fastq(fastq):

        match = parser_re.search(read).groupdict()

        cb = match['CB']

        umi = match['MB']

        cb_counter[cb] += 1

        if umi_histogram:

            umi = match['MB']

            umi_counter[(cb, umi)] += 1

    for bc, count in cb_counter.most_common():

    ''' Filters reads with non-matching barcodes

    Expects formatted fastq files.

    '''

    with open_gzipsafe(bc1) as bc1_fh:

        bc1 = set(cb.strip() for cb in bc1_fh)

    if bc2:

        with open_gzipsafe(bc2) as bc2_fh:

            bc2 = set(cb.strip() for cb in bc2_fh)

@click.option('--out_dir', default=".")

@click.option('--readnumber', default="")

@click.option('--prefix', default="")

def demultiplex_cells(fastq, out_dir, readnumber, prefix=""):

@click.option('--cb_histogram', default=None)

@click.option('--cb_cutoff', default=0)

def demultiplex_cells(fastq, out_dir, readnumber, prefix, cb_histogram,

                      cb_cutoff):

    ''' Demultiplex a fastqtransformed FASTQ file into a FASTQ file for

    each cell.

    '''

    parser_re = re.compile(re_string)

    readstring = "" if not readnumber else "_R{}".format(readnumber)

    filestring = "{prefix}{sample}{readstring}.fq"

    cb_set = set()

    if cb_histogram:

        cb_set = get_cb_depth_set(cb_histogram, cb_cutoff)

    sample_set = set()

    batch = collections.defaultdict(list)

    parsed = 0

        parsed += 1

        match = parser_re.search(read).groupdict()

        sample = match['CB']

        if cb_set and sample not in cb_set:

            continue

        sample_set.add(sample)

        batch[sample].append(read)

        # write in batches to avoid opening up file handles repeatedly