Use pLDDT values from alphafold2 to define rigid domains

[dsclassen]

@michalhammel and I have this idea to use the pLDDT values (AF2 FAQ) from an AlphaFold model to automagically define Rigid Domains in a protein. This a cool idea, but there are a few potential problems

• CHARMM-GUI currently changes all B-factors in output *.crd file to 0.00
• Would need User to upload original PDB file along with the CRD file
• The chain IDs would not line up because CHARMM-GUI uses internal segid format PRO[A-Z] (protein), DNA[A-Z] (DNA), RNA[A-Z] (RNA), and HET[A-Z] (ligands), instead of PDB chain id.
• It is not obvious to me how we would assign Rigid Domains to different Rigid Bodies. Maybe use the PAE data?

[dsclassen]

AlphaFold produces a per-residue estimate of its confidence on a scale from 0 - 100. This confidence measure is called pLDDT and corresponds to the model’s predicted score on the lDDT-Cα metric. It is stored in the B-factor fields of the mmCIF and PDB files available for download (although unlike a B-factor, higher pLDDT is better). pLDDT is also used to colour-code the residues of the model in the 3D structure viewer. The following rules of thumb provide guidance on the expected reliability of a given region:

• Regions with pLDDT > 90 are expected to be modelled to high accuracy. These should be suitable for any application that benefits from high accuracy (e.g. characterising binding sites).
• Regions with pLDDT between 70 and 90 are expected to be modelled well (a generally good backbone prediction).
• Regions with pLDDT between 50 and 70 are low confidence and should be treated with caution.
• The 3D coordinates of regions with pLDDT < 50 often have a ribbon-like appearance and should not be interpreted. We show in our paper that pLDDT < 50 is a reasonably strong predictor of disorder, i.e. it suggests such a region is either unstructured in physiological conditions or only structured as part of a complex. (Note: this relationship has typically been tested in the context of well-studied proteins, which may have more evolutionarily-related sequences available than a randomly chosen UniProt entry.)
• Structured domains with many inter-residue contacts are likely to be more reliable than extended linkers or isolated long helices.
• Unphysical bond lengths and clashes do not usually appear in confident regions. Any part of a structure with several of these should be disregarded.

Note that the PDB and mmCIF files contain coordinates for all regions, regardless of their pLDDT score. It is up to the user to interpret the model judiciously, in accordance with the guidance above.

[dsclassen]

Phenix has a tool (process_predicted_model) which we might find helpful for this task

[dsclassen]

I'm going to close this for now. As of v0.0.11 we have added a first draft of a simple Jiffy which allows users to upload their coordinates as a CHARMM *.crd file and the AlphFold PAE results as a *.json file. We then run Michal's write_const_from_pae.py script on the backend to automagically define rigid and flexible domains and create a const.inp file for download.