WatCon.find_conserved_networks module

Cluster water coordinates and analyze conservation to clustered networks

WatCon.find_conserved_networks.cluster_coordinates_only(coordinate_list, cluster='hdbscan', min_samples=10, eps=0.0, n_jobs=1)[source]

Cluster a set of coordinates.

Parameters:

  • coordinate_list (list of array-like) – A combined list of all coordinates to be clustered.

  • cluster (str) – Clustering method, can be ‘optics’, ‘dbscan’, or ‘hdbscan’.

  • min_samples (int) – Minimum number of samples required for a cluster.

Returns:

  • Cluster labels (array-like)

  • Cluster centers (dict)

Return type:

tuple

WatCon.find_conserved_networks.cluster_nodes(combined_graph, cluster='hdbscan', min_samples=10)[source]

Cluster node positions from a combined NetworkX graph.

Parameters:

  • combined_graph (networkx.Graph) – A combined graph used for clustering.

  • cluster (str) – Clustering method, can be ‘optics’, ‘dbscan’, or ‘hdbscan’.

  • min_samples (int) – Minimum number of samples required for a cluster.

Returns:

  • Cluster labels (array-like)

  • Cluster centers (dict)

Return type:

tuple

WatCon.find_conserved_networks.collect_coordinates(pkl_list)[source]

Collect coordinates into one array from .pkl files

Parameters:

pkl_list (list) – List of pkl_files (full paths)

Returns:

Array of combined coordinates

Return type:

np.ndarray

WatCon.find_conserved_networks.combine_graphs(list_of_graphs)[source]

Combine multiple NetworkX graph objects into a single graph.

Parameters:

list_of_graphs (list of networkx.Graph) – List of NetworkX graph objects to be merged.

Returns:

A combined graph containing all nodes and edges from input graphs.

Return type:

networkx.Graph

WatCon.find_conserved_networks.create_clustered_network(clusters, max_connection_distance, create_graph=True)[source]

Create a WaterNetwork object from cluster centers.

Parameters:

  • clusters (array-like) – List of XYZ coordinates of cluster centers.

  • max_connection_distance (float) – Maximum allowed distance between two clusters to form an interaction.

  • create_graph (bool, optional) – Whether to create a NetworkX graph from the clustered WaterNetwork. Default is True.

Returns:

A WaterNetwork object representing the clustered water network.

Return type:

WaterNetwork

WatCon.find_conserved_networks.find_clusters_from_densities(density_file, output_name=None, threshold=1.5)[source]

Find clusters from densities.

Parameters:

  • density_file (str) – .dx file containing density information

  • output_name (str) – Base name for output

  • threshold (float) – Threshold for cutoff

Returns:

Array of density hotspot coordinate locations

Return type:

np.ndarray

WatCon.find_conserved_networks.find_commonality(networks, centers, names, dist_cutoff=1.5, local_dens_radius=6)[source]

Find the commonality of a list of networks relative to a summary network created from clustering.

Parameters:

  • networks (list of WaterNetwork) – List of WaterNetwork objects to be analyzed.

  • centers (array-like) – Locations of clustered centers.

  • names (list[str]) – List of IDs

  • dist_cutoff (float) – Distance cutoff for classifying conserved water

  • local_dens_radius (float) – Radius cutoff for calculating local water density

Returns:

A dictionary containing calculated commonalities for each network.

Return type:

dict

WatCon.find_conserved_networks.get_coordinates_from_pdb(pdb_file)[source]

Returns coordinates for all “ATOM” or “HETATM” lines – useful for getting coordinates from cluster PDBS.

Parameters:

pdb_file (str) – Full path to PDB file.

Returns:

Array of coordinates

Return type:

np.ndarray

WatCon.find_conserved_networks.get_coordinates_from_topology(pdb_file, atom_selection='all')[source]

Collect coordinates from a given MDAnalysis-readable topology file

Parameters:

pdb_file (str) – Full path to MDAnalysis-readable topology file

Returns:

Array of coordinates

Return type:

np.ndarray

WatCon.find_conserved_networks.identify_clustered_angles(classification_file)[source]

Cluster two-angle calculations and count frequencies of clusters. Finds the closest point to the center of the cluster.

Parameters:

classification_file (str) – Path to angle classification file (.csv)

Returns:

cluster_conservation_dict – Dictionary containing, for each residue, frequency of individual clusters and water/protein atom coords corresponding (as close as possible) to each cluster

Return type:

dict

WatCon.find_conserved_networks.identify_conserved_water_clusters(networks, centers, dist_cutoff=1.0, filename_base='CLUSTERS')[source]

Create a dictionary of cluster conservation and generate a PDB file of clusters.

Parameters:

  • networks (list of WaterNetwork) – List of WaterNetwork objects to analyze.

  • centers (array-like) – List of XYZ coordinates representing cluster centers.

  • dist_cutoff (float) – Distance cutoff to classify a water molecule as part of a cluster.

  • filename_base (str) – Base filename for saving projected clusters.

Returns:

A dictionary mapping cluster centers to the count of included waters.

Return type:

dict

WatCon.find_conserved_networks.identify_conserved_water_interactions_clustering(networks, clusters, max_connection_distance=2.0, dist_cutoff=1.0, filename_base='CLUSTER', out_dir='cluster_pdbs')[source]

Rank water-water interactions in relation to clustering.

Parameters:

  • networks (list of WaterNetwork) – List of WaterNetwork objects to be analyzed.

  • centers (array-like) – List of XYZ coordinates of cluster centers.

  • max_connection_distance (float) – Maximum allowed distance between two clusters to form an interaction.

  • dist_cutoff (float) – Distance cutoff to classify a water molecule as part of a cluster.

  • filename_base (str) – Base filename for saving projected clusters.

Returns:

A dictionary mapping cluster interactions to the count of included waters.

Return type:

dict

WatCon.find_conserved_networks.plot_commonality(files=None, input_directory=None, cluster_pdb=None, commonality_dict=None, plot_type='bar', output='commonality', out_dir='images')[source]

Plot commonality to cluster centers.

Parameters:

  • files (list) – List of outputted .pkl files from WatCon

  • input_directory (str) – Directory containing files

  • cluster_pdb (str) – PDB of clusters

  • commonality_dict (dict) – Commonality dict

  • plot_type ({‘bar’, ‘hist’}) – Type of plot

  • output (str) – Base filename for saving images

Return type:

None