conversion_utils

Auxiliary functions for the conversion module.

Functions

move_topology_pdb(topology_name, origin_dir, ...)

Move a .pdb file from origin to destination directory, to act as a topology file.

join_pdbs(pdbs_dir, multimodel_name, multimodel_dir[, ...])

Join a randomly sampled number of .pdb files in a directory into a single multimodel .pdb

calc_saxs_data(trajectory_id, universe, frame_index, ...)

Back-calculate a SAXS profile for a frame of a Universe object using Pepsi-SAXS.

calc_chi2_fit(exp, calc[, sample_weights])

Apply a chi-square goodness-of-fit test between experimental and calculated data profiles.

Module Contents

ensemblify.conversion.conversion_utils.move_topology_pdb(topology_name, origin_dir, destination_dir)

Move a .pdb file from origin to destination directory, to act as a topology file.

Given a path to a directory containing .pdb files, moves a single .pdb from that directory to a destination directory, to serve as a topology file for future trajectory analysis.

Parameters:

  • topology_name (str) – Prefix identifier for moved .pdb file.

  • origin_dir (str) – Directory where file of interest is located.

  • destination_dir (str) – Directory where file will be moved to.

Returns:

Path to the moved topology file.

Return type:

str

ensemblify.conversion.conversion_utils.join_pdbs(pdbs_dir, multimodel_name, multimodel_dir, n_models=None, topology_path=None)

Join a randomly sampled number of .pdb files in a directory into a single multimodel .pdb file.

Parameters:

  • pdbs_dir (str) – Path to directory where numbered .pdb files are stored.

  • multimodel_name (str) – Prefix identifier for created multimodel .pdb file.

  • multimodel_dir (str) – Path to directory where ensemble .pdb file will be created.

  • n_models (int, optional) – Number of .pdb files to randomly sample from the specified directory. If None, all .pdb files in the directory will be used.

  • topology_path (str, optional) – Path to a topology .pdb file to be ignored when sampling .pdb files from multimodel_dir. Required if pdbs_dir matches multimodel_dir. Defaults to None.

Returns:

Path to created multimodel ensemble .pdb file.

Return type:

str

ensemblify.conversion.conversion_utils.calc_saxs_data(trajectory_id, universe, frame_index, exp_saxs_file, calc_saxs_log, pepsi_saxs_opt=None)

Back-calculate a SAXS profile for a frame of a Universe object using Pepsi-SAXS.

Calculation is done for a single temporary .pdb file created from the current frame of the MDAnalysis Universe object.

Parameters:

  • trajectory_id (str) – Prefix identifier for created files.

  • universe (mda.Universe) – Universe object containing the trajectory being analyzed.

  • frame_index (int) – Current frame being calculated.

  • exp_saxs_file (str) – Path to .dat file with experimental SAXS data for the current protein to be used by Pepsi-SAXS.

  • calc_saxs_log (str) – Path to .log file for the SAXS profile calculation of each frame.

  • pepsi_saxs_opt (str, optional) – This string will be passed onto Pepsi-SAXS as additional command line options. If None, default Pepsi-SAXS options are used instead.

Returns:

Array with the values for the SAXS profile calculated from this frame of the trajectory in the Universe object.

Return type:

np.ndarray

ensemblify.conversion.conversion_utils.calc_chi2_fit(exp, calc, sample_weights=None)

Apply a chi-square goodness-of-fit test between experimental and calculated data profiles.

Parameters:

  • exp (np.ndarray) – Experimental data in the format {value, error}.

  • calc (np.ndarray) – Calculated samples for each data point, in the format {value}. Will be averaged across samples before fitting.

  • sample_weights (np.ndarray, optional) – Weights for each sample in the calculated data. If None, all samples are assumed to have equal weight. Defaults to None.

Returns:

float:

Reduced chi2 value of fit.

np.ndarray:

Residuals of the fit.

Return type:

tuple[float, np.ndarray]