schrodinger.application.desmond.kinetics.utils module¶

class schrodinger.application.desmond.kinetics.utils.RmsdSelection(aligned_weights, displaced_weights, atom_idcs)¶

Bases: tuple

aligned_weights: List[float]¶: Alias for field number 0

displaced_weights: List[float]¶: Alias for field number 1

atom_idcs: List[int]¶: Alias for field number 2

__contains__(key, /)¶: Return key in self.

__len__()¶: Return len(self).

count(value, /)¶: Return number of occurrences of value.

index(value, start=0, stop=9223372036854775807, /)¶

Return first index of value.

Raises ValueError if the value is not present.

class schrodinger.application.desmond.kinetics.utils.MtdInfo(value)¶

Bases: enum.Enum

An enumeration.

KILL_NOW = 'kill_now'¶

LAST_INDEX = 'last_index'¶

MINDIST = 'mindist'¶

TAU_ESTIMATE_SEC = 'tau_estimate_sec'¶

SIM_TIME = 'sim_time'¶

TOO_SLOW = 'too_slow'¶

STATUS = 'status'¶

class schrodinger.application.desmond.kinetics.utils.MtdStatus(value)¶

Bases: enum.Enum

The Status of each MtD unbinding job

OK = 'OK'¶

TOO_SLOW = 'Too Slow'¶

BOUND = 'Bound'¶

schrodinger.application.desmond.kinetics.utils.find_ligand_sites(st: schrodinger.structure._structure.Structure, asl: str, nsites: int, sampling_size: int = 5000) → List[int]¶: Find a nsites atoms which are furthest away from each other on a provided small molecule. These atoms (sites) are then used in distance calculations.

schrodinger.application.desmond.kinetics.utils.find_protein_sites(st: schrodinger.structure._structure.Structure, asl: str, area_cutoff: float = 40.0) → List[int]¶

Find solvent exposed protein residues which will be used to check if the ligand is still bound to the protein.

Returns: a list of C-alpha atoms indices of such residues.

schrodinger.application.desmond.kinetics.utils.plot_descriptors(rmsd: numpy.array, mindist: numpy.array, endpoints: Tuple[Optional[int], Optional[int]], rmsd_limits: Tuple[float, float], mindist_limits: Tuple[float, float])¶: Plot descriptors from a RAMD trajectory

schrodinger.application.desmond.kinetics.utils.extract_reactive_path(cms_filename: str, asl_receptor: str, asl_ligand: str, rmsd_on: float, rmsd_off: float, mindist_on: float, mindist_off: float, strict: bool, pad_traj_start: int = 0, max_traj_frames: int = 2000, debug: bool = False) → List[str]¶

Reactive path is equivelent is a the unbinding path of the ligand

Parameters

rmsd_on – value to start recording ligand detachment cutoff
rmsd_off – value to stop recording ligand detachment cutoff, this value should be larger than rmsd_on
mindist_on – minimal distance between receptor and ligand to start recording ligand detachment cutoff
mindist_off – minimal distance between receptor and ligand to stop recording ligand detachment cutoff. This value should be larger than rmsd_on
strict – mindist_off and rmsd_off must be both satisfied for reactive trajectory extraction
pad_traj_start – number of initial trajectory frames to prepend to the reactive trajectory frames
max_traj_frames – maximum number of frames to load. If trajectory contains more frames than this value, then we will slice the trajectory such that the specified number of frames is used in the downstream analysis

Returns

a list of filenames to be saved in the analsysis stage of the workflow

schrodinger.application.desmond.kinetics.utils.calc_tension(rmsd_matrix: numpy.array, ivec: Optional[List[int]] = None) → float¶

schrodinger.application.desmond.kinetics.utils.path_rmsd_matrix(sts: List[schrodinger.structure._structure.Structure], rmsd_atoms: List[int]) → numpy.array¶: calculate 2d RMSD of the ligands on prealigned structures

schrodinger.application.desmond.kinetics.utils.path_optimize(current_guess: List[int], rmsd_matrix: numpy.array, step: float) → List[int]¶: Provided that you already have a rough path (init.mae) you can now optimize it, whiere the best set of available frames can be used. :return ” a list with frame IDs, which will be used to extract the path

schrodinger.application.desmond.kinetics.utils.path_satisfies_mindist(sts: List[schrodinger.structure._structure.Structure], asl_receptor: str, asl_ligand: str, mindist_off: float) → bool¶: Evaluate if the provided path satisfies the mindist criteria

schrodinger.application.desmond.kinetics.utils.trj2sts(cms_model: cms.Cms, tr: List[traj.Frame], path_frames: Optional[List[int]] = None) → List[schrodinger.structure._structure.Structure]¶: Convert trajectory frames to to structure objects. path_frames is a list of frames to convert to structures.

schrodinger.application.desmond.kinetics.utils.path_smooth(sts: List[schrodinger.structure._structure.Structure], atoms_receptor: List[int], atoms_ligand: List[int]) → List[schrodinger.structure._structure.Structure]¶: As final path optimization step, we will shuffle each waypoint towards or away from its neightbors such that RMSD between all the waypoints is equal. This approach is using a string-like method to optimize the distances between each waypoint. :return : a new optimized path

schrodinger.application.desmond.kinetics.utils.reduce_and_optimize_path(asl_receptor: str, asl_ligand: str, nwaypoints: int, mindist_off: float, debug: bool = False) → List[str]¶

Parameters: nwaypoints – the number of snapshots/waypoints in the unbinding path to generate
Returns: a list of filenames to be saved in the analysis stage

schrodinger.application.desmond.kinetics.utils.setup_ramd_system(asl_receptor: str, asl_ligand: str, cms_filename: str, ligand_rmsd_cutoff: float, debug: bool = False) → Optional[Cms]¶

Measure Ligand RMSD and determine if its suitable for downstream RAMD Calculation. Setup subsequent RAMD systems.

Returns: CMS file if the ligand does not unbinds during the relaxation stage (ligand RMSD is less than ligand_rmsd_cutoff). If it does unbinds then we return None.

schrodinger.application.desmond.kinetics.utils.cluster_unbinding_paths(paths: List[List[schrodinger.structure._structure.Structure]], asl_receptor: str, asl_ligand: str, use_weights: bool, max_iter: int) → Tuple[List[Tuple[int, int, int]], List[int]]¶

Given a list of paths, cluster them using AffinityPropogation methon. Return information about the clusters as well as their centroids.

Returns: for each cluster return a (size, centroid, label) tuple and a “paths label” path label is a cluster all paths belong in.

schrodinger.application.desmond.kinetics.utils.get_unbinding_pathway(paths_labels: List[int], cluster_info: List[Tuple[int, int, int]], unbound_labels: Optional[List[bool]] = None) → List[List[schrodinger.structure._structure.Structure]]¶

Using the clustering information, write the centroid unbinding paths. If the first N clusters have the the same number of members then we will return N centroids for these clusters.

Parameters

cluster_info – a list that contatins (size, center, label) of clusters
unbound_labels – use these labels to overwrite the centroids to to return. Here we will suppliment centroid information to prioritize unbound paths that meet MinDist criteria (MINDIST_GOOD_PROP in the first waypoint of the path should be set to True)

schrodinger.application.desmond.kinetics.utils.calculate_path_msd(path_sts: List[schrodinger.structure._structure.Structure], align_asl: str, displace_asl: str) → float¶

Function to calculate ‘lambda’ variable as described in the reference below. This variable is comparable to the inverse of mean square displacement between neigboring waypoints. See for details: -Branduardi D, et. al., JChemPhys 126(5), 054103; doi:10.1063/1.2432340

Parameters: path_sts – structrures defining unbinding path

schrodinger.application.desmond.kinetics.utils.process_atom_selection(path_sts: List[schrodinger.structure._structure.Structure], align_asl: str, displace_asl: str) → schrodinger.application.desmond.kinetics.utils.RmsdSelection¶

This function generates three atom lists used with required weights for M-expression RMSD calculation. The length of each list should be the same. Here we perform the bookkeeping to ensure the the length of these list are the same.

Parameters

path_sts – A list of structures that define unbinding path. These structures should not have any hydrogen atoms.
align_asl – Selection to align the system on
displace_asl – Selection to measure RMSD on

Returns

A tuple of three lists, all the same sizes, with weights for RMSD alignment/measurement for the first to variables, and atom indices that correspond to atoms used in RMSD calculation

schrodinger.application.desmond.kinetics.utils.insert_path_into_cms(cms_fname: str, path_sts_fn: str, combined_cms_fname: str)¶

This function reads in cms and path_structures and combines the two by inserting the coordinates for all the path waypoint (wp) structures as atom properties in full-system CT of the cms file. The path’s waypoint coordinates are stored in the properties like this: * r_tau_path_waypt00_{x,y,z} Which corresponds to the xyz coordinates in the first waypoint for that atom.

This is done due to a limitation that MSJ workflow can pass only one structure as an input.

Note: Hydrogen coordinates are not stored.

schrodinger.application.desmond.kinetics.utils.extract_path_from_cms(cms_fname: str) → Tuple[List[schrodinger.structure._structure.Structure], cms.Cms]¶

This function reads a cms file with stored path waypoints (see insert_path_in_cms()) in atom properties and extracts its coordinates into separate structures.

Parameters: cms_fname – filename to the CMS file with the waypoint coordinates

schrodinger.application.desmond.kinetics.utils.get_mtd_results(cvseq_filename: str) → pandas.core.frame.DataFrame¶: Read in cvseq data from a filename and then convert it to a DataFrame. Since the results of mtd can get quite large we are going to change precision and/or change data types of several fields, for more efficient storage.

schrodinger.application.desmond.kinetics.utils.get_mtd_info(df: pandas.core.frame.DataFrame, residence_time_cutoff: Optional[float]) → dict¶

Return a dict with key info about the MtD data in the provided DataFrame.

Parameters

df – dataframe with all the unbinding descriptors
residence_time_cutoff – a cutoff value to determine if the is too slow to unbind, in which case the workflow is terminated early

Returns

a dictionary with various descriptors calculated from the job. The: keys and their values of this dict are the following.

MtdInfo.TAU_ESTIMATE_SEC: Predicted resedence time (tau) in Seconds

MtdInfo.SIM_TIME: Total simulation time (in nanoseconds).

MtdInfo.MINDIST: Distance between the receptor and the ligand when the: simulation was terminated
MtdInfo.TOO_SLOW: A bool indicating if the unbinding event had: occured or not. A value of True suggests that no unbinding happend.

MtdInfo.STATUS: A string of the final status: ‘OK’, ‘BOUND’ ‘TOO SLOW’

schrodinger.application.desmond.kinetics.utils.calc_mean_tau_and_pval(res_times: List[float], ntrials: int = 100) → Tuple[schrodinger.application.desmond.measurement.Measurement, schrodinger.application.desmond.measurement.Measurement]¶: Return mean Tau and P-val values as a Measurement object

schrodinger.application.desmond.kinetics.utils.plot_tau_fitting(res_times: List[float], tau: schrodinger.application.desmond.measurement.Measurement, pvalue: schrodinger.application.desmond.measurement.Measurement, jobname: str)¶

Plot predicted residence time (tau) from the replicas and the fitted value. The plot will be written to the current directory with the jobname as a filename.

Parameters

res_times – Residence times from previous subjobs
tau – Fitted residence time for the ligand
pvalue – P-value for the fitted residence time
jobname – Master jobname, will be used to save plot in PNG format

schrodinger.application.desmond.kinetics.utils.plot_mtd_report(results: pandas.core.frame.DataFrame, jobname: str)¶: Generate a series of plots to expore MtD unbinding simulations.