Measures

measuring characteristics

Functions to measure characteristics of any BiobOx object

ccs(M, use_lib=True, impact_path='', impact_options='-Octree -nRuns 32 -cMode sem -convergence 0.01', pdbname='', tjm_scale=False, proberad=1.0)[source]

compute CCS calling either impact.

Parameters

  • M – any biobox object

  • use_lib – if true, impact library will be used, if false a system call to impact executable will be performed instead

  • impact_path – by default, the environment variable IMPACTPATH is sought. This allows redirecting to a specific impact root folder.

  • impact_options – flags to be passes to impact executable

  • pdbname – if a file has been already written, impact can be asked to analyze it

  • tjm_scale – if True, CCS value calculated with PA method is scaled to better match trajectory method.

  • proberad – radius of probe. Do find out if your impact library already adds this value by default or not (old ones do)!

Returns

CCS value in A^2. Error return: -1 = input filename not found, -2 = unknown code for CCS calculation

-3 CCS calculator failed, -4 = parsing of CCS calculation results failed

rgyr(M)[source]

compute radius of gyration.

Parameters

M – any biobox object

Returns

radius of gyration

sasa(M, targets=[], probe=1.4, n_sphere_point=960, threshold=0.05)[source]

compute the accessible surface area using the Shrake-Rupley algorithm (“rolling ball method”)

Parameters

  • M – any biobox object

  • targets – indices to be used for surface estimation. By default, all indices are kept into account.

  • probe – radius of the “rolling ball”

  • n_sphere_point – number of mesh points per atom

  • threshold – fraction of points in sphere, above which structure points are considered as exposed

Returns

accessible surface area in A^2

Returns

mesh numpy array containing the found points forming the accessible surface mesh

Returns

IDs of surface points

saxs(M, crysol_path='', crysol_options='-lm 20 -ns 500', pdbname='')[source]

compute SAXS curve using crysol (from ATSAS suite)

Parameters

  • M – any biobox object

  • crysol_path – path to crysol executable. If not provided, the environment variable ATSASPATH is sought instead. This allows redirecting to a specific ATSAS bin folder.

  • crysol_options – flags to be passes to impact executable

  • pdbname – if a file has been already written, crysol can be asked to analyze it

Returns

SAXS curve (nx2 numpy array)

measuring distances

Helper functions and classes are available to assess points’ pairwise interactions between points. These are divided in two categories:

  • straight line (i.e. euclidean distances, van der Waals interactions).

  • paths (e.g. used to model cross-linking data)

Helper functions for the calculation of pairwise interactions

distance_matrix(points1, points2)[source]

compute full distance matrix between two points clouds

Parameters

  • points1 – nx3 numpy array containing points coordinates

  • points2 – mx3 numpy array containing points coordinates

Returns

nxm distance matrix (numpy array)

get_neighbors(dist, cutoff)[source]

detect interfacing points (couples at less than a certain cutoff distance)

Parameters

  • dist – distance matrix

  • cutoff – maximal distance at which lennard-jones interaction is considered

Returns

list of interacting couples

lennard_jones(points1, points2, epsilon=1.0, sigma=2.7, cutoff=12.0, coeff1=9, coeff2=6)[source]

Compute m-n lennard-jones potential between two ensembles of points.

Parameters

  • points1 – nx3 numpy array containing points coordinates

  • points2 – mx3 numpy array containing points coordinates

  • epsilon – epsilon parameter

  • sigma – sigma parameter

  • cutoff – maximal distance at which lennard-jones interaction is considered

  • coeff1 – coefficient of repulsive term (m coefficient)

  • coeff2 – coefficient of attractive term (n coefficient)

Returns

m-n lennard-jones potential (kJ/mol)

class Path(points)[source]

Methods for finding shortest paths between points in a point cloud (typically a Structure object).

Two algorithms are implemented: A* and Theta*. In order to deal with lists of links within the same dataset, two methodologies are offered, global or local.

  • global builds a grid around the whole ensemble of points once, and every subsequent distance measure is performed on the same grid.

  • local a local grid is displaced to surround the points to be linked.

Global take longer to initialize, and is more memory intensive, but subsequent measures are quick. Local takes less memory, but individual measures take more time to be performed.

An accessibility graph must be initially created around the provided points cloud. This is a mesh where none of its nodes clashes with any of the provided point in the cloud. After instantiation, setup_local_search or setup_global_search must first be called (depending on whether one wants to generate a single grid encompassing all the protein, or a smaller, moving grid).

Parameters

points – points for representing obstacles.

a_star(start, goal)[source]

A* algorithm, find path connecting two points in the graph.

Parameters

  • start – starting point (flattened coordinate of a graph grid point).

  • goal – end point (flattened coordinate of a graph grid point).

lazy_theta_star(start, goal)[source]

Lazy Theta* algorithm (better than Theta* in terms of amount of line-of-sight tests), find path connecting two points in the graph.

Parameters

  • start – starting point (flattened coordinate of a graph grid point).

  • goal – end point (flattened coordinate of a graph grid point).

search_path(start, end, method='theta', get_path=True, update_grid=True, test_los=True)[source]

Find the shortest accessible path between two points.

Parameters

  • start – coordinates of starting point

  • end – coordinates of target point

  • method – can be theta or astar

  • get_path – return full path (not only waypoints)

  • update_grid – if True, grid will be recalculated (for local search only)

  • test_los – if true, a line of sight postprocessing will be performed to make paths straighter

setup Path to perform path search using the a global grid wrapping all the obtacles region. This method (or setup_local_search) must be called before and path detection can be lauched with search_path.

Parameters

  • step – grid step size.

  • maxdist – clash detection threshold

  • use_hull – if True, points not laying within the convex hull wrapping around obtacles will be excluded

  • boundaries – build a grid within the desired box boundaries (if defined, maxdist parameter is ignored)

  • cloud – build a grid using a points cloud as extrema for the construction of the box. If defined, maxdist and boundaries parameters are ignored.

setup Path to perform path search using the local grid method. This method (or setup_global_search) must be called before and path detection can be launched with search_path.

Parameters

  • step – grid step size

  • maxdist – clash detection threshold

smooth(chain, move_angle_thresh=0.0)[source]

Utility method aimed at smoothing a chain produced by A* or Theta*, to make it less angular.

Parameters

  • chain – numpy array containing the list of points composing the path

  • move_angle_thresh – if angle between three consecutive points is greater than this threshold, smoothing is performed

Returns

smoothed chain (3xN numpy array)

theta_star(start, goal)[source]

Theta* algorithm, find path connecting two points in the accessibility graph.

Parameters

  • start – starting point (flattened coordinate of a graph grid point).

  • goal – end point (flattened coordinate of a graph grid point).

write_grid(filename='grid.pdb')[source]

Write the accessibility graph to a PBB file

Parameters

filename – output file name

subclass of Path, measure cross-linking distance between atom pairs in a molecule.

  • after instantiation, call first set_clashing_atoms to define molecule’s atoms of interest for clash detection.

  • Subsequently, call either setup_local_search or setup_local_search to prepare the points grid used for path detection.

  • Physical distances between a list of atom indices can be finally computed with distance_matrix or, between two atoms only, with search_path.

If molecule contains multiple conformations, conformation i can be chosen by calling Xlink.molecule.set_current(i) before performing the procedure described in superclass.

Parameters

molecule – Molecule instance

distance_matrix(indices, method='theta', get_path=False, smooth=True, verbose=False, test_los=True, flexible_sidechain=False, sphere_pts_surf=4.0)[source]

compute distance matrix between provided indices.

Parameters

  • indices – atoms indices (within the data structure, not the original pdb file). Get the indices via molecule.atomselect(…) command.

  • method – can be “theta” or “astar”.

  • get_path – if true, a list containing all the paths is also returned

  • smooth – if True, path will be refined to make turns less angular.

  • verbose – if True, the algorithm will dump text in console

  • sphere_pts_surf – surface occupied per sphere point, in A2. The smaller, the higher the points density

  • test_los – if true, a line of sight postprocessing will be performed to make paths straighter

  • flexible_sidechain – if True, the selected atoms will be rotated around their associated CA, in order to scan for alternative sidechain arrangements. A sphere of clash-free alternative conformations is generated, and the shortest distance accounting for all these different possibilities is returned. Note that this method is computationally expensive.

Returns

distance matrix (numpy 2d array). matrix will contain -1 if atoms are too far, and -2 if one of the two atoms is buried. If get_path is True, a list of paths is also returned (format: [[id1, id2], [path]]).

set_clashing_atoms(atoms=[], densify=True, atoms_vdw=False, points=[])[source]

define atoms to consider for clash detection.

Parameters

  • atoms – atomnames to consider for clash detection. If undefined, protein backbone and C beta will be considered.

  • densify – if True, all atoms not solvent exposed will be considered for clash detection.

setup Path to perform path search using the a global grid wrapping all the obtacles region. This method (or setup_local_search) must be called before and path detection can be lauched with search_path.

Parameters

  • step – grid step size.

  • maxdist – clash detection threshold

  • use_hull – if True, points not laying within the convex hull wrapping around obtacles will be excluded

  • boundaries – build a grid within the desired box boundaries (if defined, maxdist parameter is ignored)

  • cloud – build a grid using a points cloud as extrema for the construction of the box. If defined, maxdist and boundaries parameters are ignored.

setup Path to perform path search using the local grid method.

This method (or setup_global_search) must be called before and path detection can be launched with search_path.

Parameters

  • step – grid step size

  • maxdist – clash detection threshold

write_protein_points(filename='protein_points.pdb')[source]

Write the points selected for clash detection into a pdb file

Parameters

filename – output file name