Pyinseq

Fixed
- Fixed [bug](https://github.com/mjmlab/pyinseq/issues/85) in generation of the `summary_gene_table.txt file`
- Fixed Travis CI link in README.md
- Added instructions for `mamba` installation, which is now required for `snakemake`
Changed
- Python version supported: 3.7, 3.8 (dropped support for 3.6)
- Updated installation instructions
Removed
- Removed conda package for pyinseq; will rely on installation via pip from PyPi or GitHub

Fixed
- File links in README are now compatible with Pypi
Added
- Pyinseq logo for documentation

Fixed
- Refactor demultiplex.py for readability and faster execution
- Logger is now compatible with snakemake logging module
- Fixed small parsing bugs in gbk_convert.py

Changed
- Remove third-party folder which contained `bowtie` executables (now `bowtie` is installed via conda or manually)
- file paths are now handled by pathlib
- New settings class that is compatible with snakemake
- The map_reads script is now a stand-alone script
- Settings class which is used by snakemake and pyinseq
- Pyinseq and snakemake now use `pathlib` for handling file paths
- Removed readthedocs website (documentation is now in the README.md; no longer using ReadTheDocs)
- Removed docs/
- Removed third-party/

Added
- Snakemake can now execute workflows using Snakefile in pyinseq
- Additional Snakefiles for genomeprep and demultiplex of pyinseq
- Conda virtual environment files that allow conda to install `bowtie` software during runtime
- Parser.py script for parsing command-line arguments
- Include '--test' option for running pytest on pyinseq
- New pytests for testing workflows in pyinseq
- Pyinseq will now summarize sample information and snakemake output
- New user guide for pyinseq

Fixed
- `pyinseq` alone brings up the help documentation
- Small fix to the `three_primeness` calculation.
A minimum of 3 reads is now required per site, and a Left:Right max read ratio of 10-fold to be tallied.

Changed
- Only Python 3.6 and 3.7 are supported.
- `screed` module is used for opening/writing fastq files.

Added
- `pyinseq genomeprep` subcommand will prepare genome files for pyinseq run. Also checks GenBank files before running.
- Added T50 calculation for sites files.
- Added progress bar for `demultiplex` function and for `writing` reads to sample files.
- `test_script.py` now compares directories and files from `pyinseq` runs to the expected output.
- Parameter `--min_counts`: minimum number of reads at a site required to be tallied. Default is 3
- Parameter `--max_ratio`: max ratio allowed between left and right reads around a TA insertion site. Default is 10-fold.
- Parameter `--transposon_seq`: define transposon sequence that is found at end of reads to help in demultiplexing. Default is ACAGGTTG
- Parameter `--barcode_length`: length of barcode index used to demultiplex reads into samples, allows for 4 - 16 nt. Default is 4.
- Parameter `--gff3`: enables `pyinseq` to write gff3 version of genome files.

Added
- Documentation hosting `http://pyinseq.readthedocs.io/`.
- `pyinseq demultiplex` command (including --notrim option).
- Improved logging of messages during the run.
- `log.txt` file to record log output to file.

Added
- Initial release.
- `pyinseq` command demultiplexes and maps samples.
- Automated testing using pytest, on TravisCI