Evcouplings

* Replaces PDB structure retrieval using the MMTF format with bCIF format (required due to MMTF server shutoff at RCSB)
* Fixes many other small incompatibilities and regressions that have accumulated over time

Update release with the following major changes - please note that you will need to update your config files due to the introduction of new mandatory parameters.

Integration of novel logistic regression-based model to assign globally comparable probabilities to ECs (probability of the EC being actually true, instead of probability of EC being in the tail of the distribution). Select using scoring_model: logistic_regression in the couplings section of your monomer config, or use "skewnormal" or "normal" to use the previously used scoring models

Improved contact map plotting: showing alignment coverage, missing structural coverage and used PDB IDs

Fixing several deprecations that led to incompatibilities with recent versions of pandas

Behind the scenes: Improved database syncing of results for EVcouplings webserver (available at http://v2.evcouplings.org)

Minor fixes and enhancements release (mostly to keep SIFTS database generation functional)

This release extends the protein complex pipeline with mutation effect prediction and docking.

This new release includes:

- Bug fixes in compare stage
- Revamped example config files
- More tests