Polyclonal

------
- Flatten mutation-escape values for RBD simulation so most values are roughly zero. Previously many values were ~0.5 rather than 0. Also adjust activities for this simulation.
- Improve column spacing when ``Polyclonal.fit`` prints log.
- Added 'atom' as output column in ``pdb_utils.extract_atom_locations``
- Added ``pdb_utils.inter_residue_distances``
- Added ``Polyclonal.distance_matrix`` attribute, set via ``spatial_distances`` parameter.
- Added spatial regularization to ``Polyclonal.fit``
- Added uniqueness regularization to ``Polyclonal.fit`` as an alternative to similarity regularization that does not go with square of site-level values.

---
- Add epitope similarity regularization that can be tuned by the parameter ``reg_similarity_weight``.
- Add ``real_mAb_cocktail.ipynb`` notebook that fits model to a real mAb cocktail dataset.

---
- Fix bug introduced in version 2.3 that dropped wildtype sites if there were minimums set in ``slider_binding_range_kwargs`` to ``lineplot_and_heatmap``.

---
- ``lineplot_and_heatmap`` computes the limit for the heatmap range **after** applying the minimum filters specified in the filters. This avoids having the range determined by mutations that are never plotted, and so is sort of a bug fix (prior behavior wasn't strictly a bug, but did not give sensible behavior).

---
- Require at least ``pandas`` 1.5.
- Some minor changes to avoid ``pandas`` warnings about future deprecations.
- For ``Polyclonal`` initialization, allow `data_mut_escape_overlap` != "exact" even with `sites` set

---
- ``lineplot_and_heatmap`` filters site max value after other slider filters, this gives correct behavior and is sort of a bug fix for the plots.
- Add ``slider_binding_range_kwargs`` to ``lineplot_and_heatmap``
- Allow ``df_to_merge`` to be list for ``mut_escape_plot`` methods.