add total frustration functions and an example notebook

[eagalpern]

Description

I added functions at the end of the frustration.py file to compute energy using the proper masks for segments "in context". I also included a function to compute total frustration. For the frustration of an "independent" fragment, the total frustration should be computed using a sliced pdb model using only that fragment.

The new functions use explicit decoy computation. Sequence decoys are created using the lenght and amino acid composition of the native sequence, and their energy is computed directly. An "approximation" for the total configurational frustration is available also.

I added the total frustration function to the Frustratometer class.
I created a new example notebook to use these new functions

Todos

• Improve the description of the new functions in the frustration.py file
• Add functions to compute the sliding window frustration (using the "in context" energy computation).

Note

The total frustration that can be directly computed using this module is not directly comparable with the PNAS 2024 results for exon relative total frustration. For that, it should be taken into account that the sequences and potts models do not cover the complete PDBs and that the native sequences used were not PDB sequences, but instead the ones having the analyzed exon.