drugforge.spectrum.calculate_rmsd.select_best_colabfold

drugforge.spectrum.calculate_rmsd.select_best_colabfold(results_dir: str, seq_name: str, pdb_ref: str, chain='A', final_pdb='aligned_protein.pdb', fold_model='alphafold2_ptm') → tuple[float, str][source]

Select the best seed output (repetition) from a ColabFold run based on its RMSD wrt the reference.

Parameters:

  • results_dir (str) – The directory containing the ColabFold results.

  • seq_name (str) – The name we gave to the sequence in the csv file.

  • pdb_ref (str) – The path to the PDB of te reference protein.

  • chain (str, optional) – Chain of both reference and generated PDB that will be used, by default “A”

  • final_pdb (str, optional) – Path to the PDB where aligned structure will be saved, by default “aligned_protein.pdb”

  • fold_model (str, optional) – The model used for ColabFold, by default “alphafold2_ptm”

Returns:

RMSD after alignment, Path to saved PDB

Return type:

Tuple[float, str]

Raises:

FileNotFoundError – The directory with ColabFold results does not exist