Prolif

Added
- `Fingerprint.run` now has a `converter_kwargs` parameter that can pass kwargs to the
underlying RDKitConverter from MDAnalysis (Issue 57).
- Formatting with `black`.

Changed
- The SMARTS for the following groups have been updated to a more accurate definition
(Issue 68, PR 73 by DrrDom, and PR 84):
- Hydrophobic: excluded F, Cl, tetracoordinated C and S, C connected to N, O or F.
- HBond donor: exclude charged O, S and charged aromatic N, only accept nitrogen
that is in valence 3 or ammonium
- HBond acceptor: exclude amides and some amines, exclude biaryl ethers and alkoxy
oxygen from esters, include some aromatic oxygen and nitrogen,
- Anion: include resonance forms of carboxylic, sulfonic and phosphorus acids,
- Cation: include amidine and guanidine,
- Metal ligand: exclude amides and some amines.
- The Pi stacking interactions have been changed for a more accurate implementation
(PR 97, PR 98).
- The Van der Waals contact has been added to the default interactions, and the `tolerance`
parameter has been set to 0.
- The `pdbqt_supplier` will not add explicit hydrogen atoms anymore, to avoid detecting
hydrogen bonds with "random" hydrogens that weren't in the PDBQT file (PR 99).
- When using the `pdbqt_supplier`, irrelevant warnings and logs have been disabled (PR 99).
- Updated the minimal RDKit version to `2021.03.1`

Fixed
- Dead link in the quickstart notebook for the MDAnalysis quickstart (PR 75, radifar).
- The `pdbqt_supplier` now correctly preserves hydrogens from the input PDBQT file (PR 99).
- If no interaction was detected, `to_dataframe` would error without giving a helpful message. It
now returns a dataframe with the correct number of frames in the index and no column.

Added
- Support for multiprocessing, enabled by default (Issue 46). The number of processes can
be controlled through `n_jobs` in `fp.run` and `fp.run_from_iterable`.
- New interaction: van der Waals contact, based on the sum of vdW radii of two atoms.
- Saving/loading the fingerprint object as a pickle with `fp.to_pickle` and
`Fingerprint.from_pickle` (Issue 40).

Changed
- Molecule suppliers can now be indexed, reused and can return their length, instead of
being single-use generators.

Fixed
- ProLIF can now be installed through pip and conda (Issue 6).
- If no interaction is detected in the first frame, `to_dataframe` will not complain about
a `KeyError` anymore (Issue 44).
- When creating a `plf.Fingerprint`, unknown interactions will no longer fail silently.

Added
- Added our J. Cheminformatics article to the citation page of the documentation and the
`CITATION.cff` file.

Changed
- Improved the documentation on how to properly restrict interactions to ignore the
protein backbone (Issue 22), how to fix the empty dataframe issue when no bond
information is present in the PDB file (Issue 15), how to save the LigNetwork diagram
(Issue 21), and some clarifications on using `fp.generate`

Fixed
- Mixing residue type with interaction type in the interactive legend of the LigNetwork
would incorrectly display/hide some residues on the canvas (PR 23)
- MOL2 files starting with a comment (``) would lead to an error

Changed
- Custom interactions must return three values: a boolean for the interaction,
and the indices of residue atoms responsible for the interaction

Fixed
- Custom interactions that only returned a single value instead of three would
raise an uninformative error message

Added
- LigNetwork: an interaction diagram with atomistic details for the ligand and
residue-level details for the protein, fully interactive in a browser/notebook, inspired
from LigPlot (PR 19)
- `fp.generate`: a method to get the IFP between two `prolif.Molecule` objects (PR 19)

Changed
- Default residue name and number: `UNK` and `0` are now the default values if `None` or
`''` is given
- The Hydrophobic interaction now uses `+0` (no charge) instead of `!$([+{1-},-{1-}])`
(not negatively or positively charged) for part of its SMARTS pattern (PR 19)
- Moved the `return_atoms` parameter from the `run` methods to `to_dataframe` to avoid
recalculating the IFP if one wants to display it with atomic details (PR 19)
- Changed the values returned by `fp.bitvector_atoms`: the atom indices have been
separated in two lists, one for the ligand and one for the protein (PR 19)

Fixed
- Residues with a resnumber of `0` are not converted to `None` anymore (Issue 13)
- Fingerprint instantiated with an unknown interaction name will now raise a `NameError`

Added
- Integration with Zenodo to automatically generate a DOI for new releases
- Citation page
- Docking section in the Quickstart notebook (Issue 11)
- PDBQT, MOL2 and SDF molecule suppliers to make it easier for users to use docking
results as input (Issue 11)
- `Molecule.from_rdkit` classmethod to easily prepare RDKit molecules for ProLIF

Changed
- The visualisation notebook now displays the protein with py3Dmol. Some examples for
creating and displaying a graph from the interaction dataframe have been added
- Updated the installation instructions to show how to install a specific release
- The previous repr method of `ResidueId` was easy to confuse with a string, especially
when trying to access the `Fingerprint.ifp` results by string. The new repr method is
now more explicit.
- Added the `Fingerprint.run_from_iterable` method, which uses the new supplier functions
to quickly generate a fingerprint.
- Sorted the output of `Fingerprint.list_available`

Fixed
- `Fingerprint.to_dataframe` is now much faster (Issue 7)
- `ResidueId.from_string` method now supports 1-letter and 2-letter codes for RNA/DNA
(Issue 8)