Fix wrong setup of LJ, improve CI and coverage

[pm-blanco]

Quite dense PR, I would appreciate at least two reviewers. Thank you @mariusaarsten for the help while addressing this PR. Partially solves #58.

Bug fixes:

• closes Fix issue on create_paper_data.py #59
• Fixes a heavy bug introduced in Refactor lj parameters #27. In that PR, we replaced diameter by sigma, but we did not correct the setup for the sample scripts with particles with different sizes. I have added a proper offset to keep the intended previous behaviur.
• Different behaviors were obtained depending on which order pmb.load_pka_set and pmb.load_interaction_parameters were called. Now one gets consistent behavior independently of the order.
• pmb.read_protein_vtf_in_df returned the estimated diameter topology_dict["radius"] instead than the radius
• Removes the Pandas warnings from lib/create_cg_from_pdb.py

Code improvements:

• I have deprecated the following protein-specific functions:

• pymbe_library.define_AA_particles_in_sequence()
• pymbe_library.define_epsilon_value_of_particles()
• pymbe_library.define_particles_parameter_from_dict()
• pymbe_library.setup_particle_sigma()
• pymbe_library.write_output_vtf_file()

• I have merged them in a more general function to define the parameters of several particles pymbe_library.define_particles()
• pymbe_library.write_output_vtf_file() was not really necesary since there already exists a built-in function in espresso that does the same job and furthermore the output vtf were not compatible with visualization/vmd-traj.py.
• I have removed the case-specific arguments from samples/Beyer2024/globular_protein.py. Now the setup of the metal ions is done automatically inside define_protein. I have also defined new functions pymbe_library.get_metal_ions_charge_map() and pymbe_library.check_if_metal_ion() to deal with the metal ions to make easier later the perspective generalization that we discussed @paobtorres @jngrad

New CI tests for the following functions (some of them very much needed):

• pymbe_library.calculate_net_charge()
• pymbe_library.create_added_salt()
• pymbe_library.create_counterions()
• pymbe_library.load_interaction_parameters()
• pymbe_library.check_aminoacid_key()
• pymbe_library.check_if_metal_ion()
• pymbe_library.get_metal_ions_charge_map()
• pymbe_library.define_particles()
• pymbe_library.define_particle()
• pymbe_library.define_residue()
• pymbe_library.define_molecule()
• pymbe_library.create_particle()
• pymbe_library.create_residue()
• pymbe_library.create_molecule()

I ran the functional tests and they all pass. I also ran longer tests and the results seem pretty similar to the results in our publication.

Current coverage: 77% total, 70% in pymbe.py

[paobtorres]

I tested every file that was changed, and I also reproduce the paper data, everything worked just fine.

[pm-blanco]

@davidbbeyer will you have time to take a look early next week? Otherwise I will simply proceed to merge this PR, this commit has been used by my students for the last two weeks and reviwed by @paobtorres so it seems quite stable 🙂

[davidbbeyer]

@pm-blanco Yes, on Monday.

[davidbbeyer]

Also looks ok for me