neurosnap.algos.alphafold module

Variables, functions, and classes associated with AlphaFold2 and whatnot.

neurosnap.algos.alphafold.score_af2m_binding(af2m_dict, binder_len, target_len=None)

Calculate binding scores from AlphaFold2 multimer prediction results.

The binder is assumed to be the first part of the sequence up to binder_len, with the target being the remainder, unless otherwise specified. Adapted from: hgbrian/biomodals

Parameters:

• af_multimer_dict – From AlphaFold2 multimer JSON file

• binder_len (int) – Length of the binder protein sequence

• target_len (Optional[int]) – Length of the target protein sequence

Returns:

• plddt_binder: Average pLDDT score for the binder.

• plddt_target: Average pLDDT score for the target.

• pae_binder: Average PAE score within the binder.

• pae_target: Average PAE score within the target.

• ipae: Average PAE score for the binder-target interaction.

Return type:

A dictionary containing the following keys