Refinegems

With this release:
- a badge for the Frontiers publication mentioning/ introducing refineGEMs,
- a How to cite section within the documentation main page and the README of this repository,
- a CITATION.cff enabling a citation button on the main page of this repository,
- some bug fixes,
- a new module called `util` and
- some computational time changes are added.

In addition, the `Pipfile`, `Pipfile.lock` & `MANIFEST.in` files are removed.
Moreover, the media definitions within the database `data. db` are updated.

The bug fixes are for the module `polish`.
The computational time changes are for the module `gapfill`.

* Made internal database in refineGEMs available to be used in user-specific scripts
* Enhanced the `get_bigg2other_db` function in the `analysis_db` module to also obtain a mapping from the ModelSEED namespace to the BiGG namespace
* Enhanced the `get_model_reacs_or_metabs` function in the `entities` module to create a pandas.DataFrame with a user-specific column name for the extracted reactions/metabolites (The default column name stayed the same to keep backwards compatibility.)

- Adjusted handling of BioCyc identifiers within the `polish_annotations` function of the module `polish`
- Added requirement for importlib_resources=5.13.0 to Pipfile
- Added code to cope with missing sub-database prefixes for BioCyc identifiers
- Changed NaN identifier handling
- Fixed issue with None prefix identifier pairs being listed in invalid_curies.tsv

The documentation displayed the functions of the polish module incorrectly. This is fixed now.

With this release, the polish module is now connected to the API of [bioregistry.io](http://bioregistry.io/).

New additions to the polish module:
+ Check for invalid CURIEs with bioregistry
+ Addition of MIRIAM compliant URIs/IRIs with bioregistry
+ Invalid URIs & CURIEs are reported back with file(s) for better manual curation
+ NaN identifiers are removed, reported with invalid URIs & CURIEs and removed
+ If 'id_db' is set to 'VMH' the BiGG identifiers retrieved from the VMH ones are checked for validity

- In the module polish bugs related to InChI-Strings in the annotations were fixed.
- The new module biomass was created to check for biomass reaction consistency and to normalise the biomass reaction if necessary on a simple level.