Pycroquet

Adds example data used in publication.

- Improve outputs when `R2_R1` has been applied to ensure input fields are replicated "as is" in counts file
- Add warning to indicate `sgrna_strands` is minimally implemented within the manifest and not applied in the processing (see 13)

- Dual-guide count output is now same as single-guide
- All count output files are compressed with gzip
- Merging of count and statistics files implemented

Use classifications in the `*.query_class.tsv.gz` file.

- Adds `mean_count_per_guide` statistic to all modes.
- Correct bug in `--boundary` mode mapping (TinQ and QinT reversed).
- dual-guide: expand classifications to handle multiple guide-pair hits when one end unique.
- dual-guide: Adds the `*.query_class.tsv.gz` file for use in debugging (see 4).

- dual-guide: added library header item to deal with reversed read order when comparing against sgRNA.
- dual-guide: handled data multiplication issue in CRAM outputs, quicker.