Umis

- Fix for kallisto output failing due to defaultdict not being imported. Thanks to andreas-wilm for the fix.
- Added `tagcount` option `--parse_tags` to use BAM tags rather than parsing read names (`UM` for UMI, `CR` for cell barcode)
- Added `tagcount` option `--gene_tags` to use BAM tags to get ID of mapping gene (`GX` tag).
- Fix tagcount with `--genemap` option not including a column name for the index.
- Add `sparse` subcommand to turn a matrix into a sparse matrix.
- Add `fasttagcount` subcommand. This assumes the input BAM/SAM file is coordinate sorted. Reduces memory usage by over
100x and runtime by 30-40% for deep samples.
- Warn, don't fail if transcripts are missing from the genemap.

- Fix skipping first piece of evidence when tagcounting.
- Add test for tagcount.
- Output full sorted transcript table from tagcount rather than only the observed transcripts.
- Add `--sparse` option to output tagcount matrices in MatrixMarket format.
- Allow cb_histogram subcommand to take gzipped files.
- Allow cb_filter subcommand to take gzipped files.
- Add support for triple-cellular barcodes.
- Add example for Illumina SureCell (https://www.illumina.com/products/by-type/sequencing-kits/library-prep-kits/surecell-wta-ddseq.html)

- Fix automatic format detection in cb_histogram.
- Add tests for cb_histogram.
- Re-enable streaming bamtagging. Thanks to chapmanb for the suggestion.
- Add subset_bamfile to subset a BAM file to keep alignments with a given set of cellular barcodes.
- Speed improvements for reading gzipped FASTQ files.
- Memory usage improvements for tagcount.

- Fix for handling unicode, thanks to chapmanb and sowmyaiyer
- Adds support for adding BAM tags to aligned fastqtransformed files. Thanks to chapmanb.
- Adds support for UMI-only fastqtransformation.
- Adds support for paired-end target sequences.
- Adds support for detecting sample barcodes via the SB tag in the regex.
- Adds support for sample-based demultiplexing with error correction.

- Now supports transforming 3-file input, as from the Linnarsson lab STRT-Seq data
- New kallisto subcommand formats read files for input to kallisto's UMI mode
- Fix gzip based fastq reading on Python 3.5
- Including preliminary subcommand for guessing cell cutoff from cb_histogram

- Added MANIFEST file which broke pip installation