Spacemake

- Fixed issues with STAR shared memory indexes

This release provides better RAM usage and faster processing through the following features:
- Lower RAM consumption during the fastq to BAM step.
- Lower RAM consumption and increased ~5x faster processing during the h5ad creation step.
- STAR now loads the genome index only once in memory and shares it across workers.

- Fixed a critical bug when processing merging samples
- Updated documentation to describe the Open-ST workflow

- Fixed several bugs.
- Added support for open-ST `run_mode`, and `barcode_flavor`.

- Updated the `environment.yaml` to pin the `pandas` version.
- Fixed bugs and errors.

- Added support for multiple references per species (`genome`, `rRNA`, `mirgenedb` etc).
- The multiple species references are controlled through the `map_strategy` variable (e.g. `rRNA->genome`).
- Implemented an optimized algorithm for binning (meshing) coordinate systems.
- Added support for easy addition of custom rules into the workflow.
- Added unit tests.
- Updated the `environment.yaml` to pin the OpenJDK version.
- Fixed several bugs and errors.