schrodinger.application.matsci.automated_cg_mapping module

Map an all-atom structure to DPD coarse-grained one

Copyright Schrodinger, LLC. All rights reserved.

schrodinger.application.matsci.automated_cg_mapping.find_new_group_id(groupids)

Find new key that is not already contained in the dict provided

Parameters

groupids (list) – the list of groupids

Return type

int

Returns

id for a new key

exception schrodinger.application.matsci.automated_cg_mapping.MappingError

Bases: Exception

Exception raised when mapping fails.

class schrodinger.application.matsci.automated_cg_mapping.CGGraphHelper

Bases: object

Includes methods to generate and analyze graphs

static classifyNodeClusters(node_clusters, st_graph)

Sort the given node clusters. Clusters that produce isomorphic subgraphs are considered equivalent, and any nodes that seed equivalent clusters are assigned to the same category together.

Parameters

  • node_clusters (dict) – dictionary of nodes_clusters where the key is the node index and value is list of cluster nodes

  • st_graph (networkx.Graph) – The structure graph

Return type

Mapping

Returns

Mapping of symmetry ids and the nodes that belong to the same category

static getPathLength(bead_nodes, st_graph)

Get the longest path length between nodes in a bead

Parameters

  • bead_nodes (list) – list of nodes in the bead

  • st_graph (networkx.Graph) – input graph

Return type

float

Returns

The longest path length between the nodes

static getCGGraph(mapping_cg, st_graph)

Get coarse-grained graph from mapping

Parameters

  • mapping_cg (Mapping) – mapping object

  • st_graph (networkx.Graph) – input full graph

Return type

networkx.Graph

Returns

the coarse grained graph

static getDualGraph(st_graph, node_clusters=None)

In a dual graph, nodes i and j are connected only if their corresponding node clusters have no overlapping nodes. If node_clusters is None, then the dual graph is constructed by connecting nodes i and j if they are at least 3 bonds away from each other.

Parameters

  • st_graph (networkx.Graph) – input graph

  • node_clusters (dict) – dictionary of node clusters for each node in the graph

  • martini_edges – list of nodes that are martini rings

Type

list or None

Return type

networkx.Graph

Returns

the dual graph

static getJoiningGraph(joinees, joiners, cg_graph, martini_edges=None)

Create a graph of the joiners and the joinees to determine which nodes should be joined. The graph is constructed by removing all edges but those between joiners and joinees. Connected components will be used to find bead pairs that will be joined.

Parameters

  • joinees (list) – list of joinee beads

  • joiners (list) – list of joiner beads

  • cg_graph (networkx.Graph) – coarse-grained graph

  • martini_edges – list of nodes that are martini rings

Type

list or None

Return type

networkx.Graph

Returns

joining graph

static getSwappingGraph(acceptors, donors, cg_graph, martini_edges=None)

Construct a graph used to determine which nodes should be swapped. Donors, the large beads will donate a node to the acceptor bead, which is a small bead. The graph is constructed by removing all edges but those between donors and acceptors. Connected components will be used to find bead pairs that will exchange nodes.

Parameters

  • acceptors (list) – list of acceptor beads

  • donors (list) – list of donor beads

  • cg_graph (networkx.Graph) – coarse-grained graph

  • martini_edges – list of nodes that are martini rings

Type

list or None

Return type

networkx.Graph

Returns

swapping graph

static modifyGraphForMartiniRings(graph, martini_edges)

Modify a the given graph by disconnecting martini rings from the rest of the graph. This should be used to prevent the rings from being swapped or joined with other beads while mapping the residue.

Parameters

  • graph (networkx.Graph) – input graph

  • martini_edges (list) – list of nodes that are martini rings

Return type

networkx.Graph

Returns

modified graph

static getEdgeList(mapping, st_graph)

Get list of edges for the coarse-grained network.

Parameters

  • mapping (Mapping) – CG mapping for a section of (or an entire) molecule

  • st_graph (networkx.Graph) – Structure graph for the underlying atomistic (not CG) representation of the molecule

Return type

tuple(list, dict)

Returns

List of all edges, and a dict containing edge attributes (bond orders for the CG bonds)

static matchIsomorphicGraphs(graphs)

Get a graph matchers object for a pair of graphs.

Parameters

graphs (tuple(Graph, Graph)) – Two graphs to be matched

Return type

iso.GraphMatcher

Returns

the object matching two graphs

static getRingDecorations(ring_nodes, res_graph)

For a given ring, find the atoms that are connected to it in the residue graph and categorize them as short (dangly) end (one atom), long (dangly) end (two atoms), and connection node (one atom) that conncets the ring to the rest of the residue.

Parameters

  • ring_nodes (list) – list of nodes in the ring

  • res_graph (networkx.Graph) – residue graph

Return type

tuple(list, list, list)

Returns

short end, long end, and connection nodes

static getWeightedRingGraph(st_graph, ring, short_end, long_end, connection_nodes)

To create a ring graph, we need to add the different weights to the dangling nodes based on thier size. The weights will be used to find centrality scores for the nodes in the ring.

Parameters

  • st_graph (networkx.Graph) – input graph

  • ring (list) – list of nodes in the ring

  • short_end (list) – list of nodes in the short end of the ring

  • long_end (list) – list of nodes in the long end of the ring

  • connection_nodes (list) – list of nodes that connect the short and long ends

Return type

networkx.Graph

Returns

the ring graph with weights for nodes

static getNodeCentralityType(st_graph)

Get a centrality type dictionary from the structure graph. This will be used to determine the type of the node based on its centrality score.

Parameters

st_graph (networkx.Graph) – input graph

Return type

dict

Returns

dictionary of node centrality type, with score as key and list of nodes as value

static hasTopologicalSymmetry(centerality_types, ring_nodes)

Decide if the structure has topological symmetry based on the centrality scores of the nodes.

Parameters

  • centerality_types (dict) – Dictionary of centrality score and the list of nodes

  • ring_nodes (list) – The nodes that are on a ring

Return type

bool

Returns

True if the structure has topological symmetry, meaning there are more more nodes with similar topology than nodes with unique topology

class schrodinger.application.matsci.automated_cg_mapping.Mapping

Bases: object

Data structure to store CG mappings of atoms into beads.

__init__()

Create an instance.

update(other_mapping)

Update the mapping and add all groups from another mapping.

Parameters

other_mapping (Mapping) – The other mapping to update with

createNewGroup(bead, bead_idx=None)

Create a new bead. If provided index is None or already exists, one is generated

Parameters

  • bead (list) – The new bead (list of nodes).

  • bead_idx (init) – The index of the new group

Return type

int

Returns

The index of the created group

addNodeToGroup(node, bead)

Add a node to a bead and if not already present and update both group_node and node_group dictionaries.

Parameters

  • node (int) – The node to be added to the bead

  • bead (int) – The index of the bead that node is being added to

removeNodeFromGroup(node, bead)

Remove a node from a bead, if present.

Parameters

  • node (int) – The node to remove

  • bead (int) – The bead that node is being removed from

deleteGroup(bead)

Delete a bead from group_node and node_group dictionaries.

Parameters

bead (int) – The bead being deleted

deleteNode(node)

Delete a node from node_group and group_node dictionaries.

Parameters

node (init) – The node being deleted

setGroup(bead_idx, bead)

Set an existing bead group.

Parameters

  • bead_idx – The bead index

  • bead (list(int)) – The bead

class schrodinger.application.matsci.automated_cg_mapping.StructureAnalyzer(struct, scale, mapping_method='dpd')

Bases: object

Analyzing input structure.

MAPPED_ATOM_INDEX = 'i_matsci_mapped_atom_index'
ORIGINAL_ATOM_INDEX = 'i_matsci_atom_index_before_extraction'
__init__(struct, scale, mapping_method='dpd')

Create an instance.

Parameters

  • struct (structure.Structure) – input structure

  • scale (int) – Coarse graining scale, number of AA atoms per CG bead

  • mapping_method (str) – Coarse graining method, MARTINI or DPD

getMolInfo()

For each unique species in the structure, create a MoleculeData object holding the species information, and store them in a list.

getUniqueSpecies()

Get unique species in the structure. Structure are found using their stereochemically unaware SMILES.

Return type

list(SpeciesData)

Returns

list of all species found in the structure

getSpeciesMolNum(mol_species)

Get a sample molecule index for the given species.

Parameters

mol_species (SpeciesData) – The species object

Return type

int

Returns

The molecule index of the species

getMappedStruct()

Map the structure by adding grouping properties to the atoms.

Return type

tuple(structure.Structure, list(str))

Returns

mapped structure and the list of CG bead names

mapTaggedAtoms()

Map the atoms that are used for mapping in sample molecules to the corresponding atoms in the rest of the structure.

addGroupingProps()

Add grouping properties to the mapped structure. The grouping properties are retrieved from sample molecules which are mapped and passed on to the rest of the structure.

removeStaleProps()

Remove properties that are no longer needed.

class schrodinger.application.matsci.automated_cg_mapping.StructData(molecule_st, scale, mapping_method='dpd')

Bases: object

A dataclass to information related to the structure of a species.

NOT_HETERO_ATOM_NUMS = (6, 14)
PREDEFINED_PATTERNS = None
ALL_SMARTS = {'Acid_Chloride': 'C(=O)Cl', 'Aldehyde': '[CH;D2;!$(C-[!#6;!#1])]=O', 'Amine': '[N;$(N-[#6]);!$(N-[!#6;!#1]);!$(N-C=[O,N,S])]', 'Azide': '[N;H0;$(N-[#6]);D2]=[N;D2]=[N;D1]', 'Boronic_Acid': '[$(B-!@[#6])](O)(O)', 'Carboxylic_acid': 'C(=O)[O;H,-]', 'Halogen': '[$([F,Cl,Br,I]-!@[#6]);!$([F,Cl,Br,I]-!@C-!@[F,Cl,Br,I]);!$([F,Cl,Br,I]-[C,S](=[O,S,N]))]', 'Isocyanate': '[$(N-!@[#6])](=!@C=!@O)', 'Nitro': '[N;H0;$(N-[#6]);D3](=[O;D1])~[O;D1]', 'Sulfonyl_Chloride': '[$(S-!@[#6])](=O)(=O)(Cl)', 'Terminal_Alkyne': '[C;$(C#[CH])]', 'dangling_atoms': '[!#1;X4&H3,X3&H2,X2&H1,X1&H0]-[!#1]', 'higher_order_bonds': '[!#1]=,#[!#1]'}
__init__(molecule_st, scale, mapping_method='dpd')

Create an instance.

Parameters

  • molecule_st (structure.Structure) – molecule structure

  • scale (int) – Coarse graining scale, number of AA atoms per CG bead

  • mapping_method (str) – Coarse graining method

getBeadSize(mapping, bead_idx)

Get the size of a bead in the provided mapping.

Parameters

  • mapping (Mapping) – mapping object

  • bead_idx (int) – index of the bead

Return type

float

Returns

size of the bead

getDpdBeadSize(mapping, bead_idx)

Get the size of a DPD bead in the provided mapping.

Parameters

  • mapping (Mapping) – mapping object

  • bead_idx (int) – index of the bead

Return type

float

Returns

size of the bead

getMartiniBeadSize(mapping, bead_idx)

Get the size of a Martini bead in the provided mapping.

Parameters

  • mapping (Mapping) – mapping object

  • bead_idx (int) – index of the bead

Return type

float

Returns

size of the bead

getNonRingContToSize(bead, mapping)

Get the contribution of non-ring nodes to the size of the bead.

Parameters

  • bead (list) – list of nodes in the bead

  • mapping (Mapping) – mapping object

Return type

float

Returns

contribution of non-ring nodes to the size of the bead

getRingNodeContToSize(bead, mapping)

Get the contribution of nodes that are on one or multiple rings to the size of the bead.

Parameters

  • bead (list) – list of nodes in the bead

  • mapping (Mapping) – mapping object

Return type

dict

Returns

contribution of nodes on the ring to the size of the bead

findExclusiveAtoms()

Find unsharable atoms in a structure.

Return type

list(int)

Returns

list of atom indexes

property sorted_func_groups

Sort functional groups based on complexity.

Return type

list(tuple(str, str))

Returns

list of functional groups sorted by complexity

createGroupMapping(struct_groups, node_to_atom_idx)

Create a mapping for all atoms that are found to belong to a defined group.

Parameters

  • struct_groups (list(tuple(str, str))) – A list of atom indexes that belong to a the definied groups.

  • node_to_atom_idx (dict) – A dictionary of node number to atom index conversion

Return type

Mapping

Returns

A mapping of functional groups

findGroupsMapping(node_to_atom_idx)

Find functional groups and pre-defined groups (if requested) in the structure.

Parameters

node_to_atom_idx (dict) – A dictionary of node number to atom index conversion

Return type

tuple(Mapping, Mapping)

Returns

A mapping of functional groups and a mapping of predefined groups. Each can be an empty mapping.

updateFunctionalGroups(func_groups_indexes, predefined_indexes)

Update functional groups by removing groups that include atoms that already belong to a predefined group. And add all predefined groups as a functional group.

Parameters

  • func_groups_indexes (list(list(int))) – list of atom indexes that match the functional groups.

  • predefined_indexes (list(list(int))) – List of atom indexes that match the predefined groups.

Return type

list(list(int))

Returns

List of atom indexes forming functional groups.

findSMARTSMatchingIndexes(smarts_info, allow_partial_overlap=True, match_scale=False)

Find indexes of atoms that match the provided smarts pattern.

Parameters

  • smarts_info (list) – SMARTS pattern and their name to be matched.

  • allow_partial_overlap (bool) – If True, allow atoms to belong to more than one group. If False, atoms can only belong to one group.

  • match_scale (bool) – If True, compare the size of the groups against the CG scale, and filter out groups that are larger than the scale.

Return type

list(list(int))

Returns

List of lists of atom indexes that match the smarts pattern.

class schrodinger.application.matsci.automated_cg_mapping.MoleculeData(molecule_st, scale, struct, mol_idx, species, mol_count, mapping_method='dpd')

Bases: schrodinger.application.matsci.automated_cg_mapping.StructData

A dataclass to hold Molecule information.

MIN_HEAVY_ATOMS = 3
UNK = 'UNK'
RESIDUE_NAME(idx)
RESIDUE

alias of schrodinger.application.matsci.automated_cg_mapping.RESIDUES

class BEAD_TYPE(bead_graph, name, beads, smarts)

Bases: tuple

bead_graph

Alias for field number 0

beads

Alias for field number 2

name

Alias for field number 1

smarts

Alias for field number 3

class NODE_GROUP_TYPE(atom_idx, bead_names, bead_nums, atom_nums)

Bases: tuple

atom_idx

Alias for field number 0

atom_nums

Alias for field number 3

bead_names

Alias for field number 1

bead_nums

Alias for field number 2

class MAPPED_BEAD_INFO(name, smarts, charge)

Bases: tuple

charge

Alias for field number 2

name

Alias for field number 0

smarts

Alias for field number 1

__init__(molecule_st, scale, struct, mol_idx, species, mol_count, mapping_method='dpd')

Create an instance.

Parameters

  • molecule_st (structure.Structure) – molecule structure

  • scale (int) – Coarse graining scale, number of AA atoms per CG bead

  • struct (structure.Structure) – The input structure

  • mol_idx (int) – molecule index in the input structure

  • species (str) – species display name

  • mol_count (int) – molecule count in the unique species

  • mapping_method (str) – Coarse graining method

processMoleculeInfo()

Process molecule information, such as type, graph and initial mapping. Find residues if the molecule is a large molecule.

getInitialMapping()

Initialize the type of molecule.

getMoleculeGraph()

Initialize molecule graph and data corresponding to molecule structure.

findResidues()

Get ResidueData objects for the sample residues found in the molecule.

getSampleResidues()

Get the unique residues in the molecule and used them as samples to be mapped. In case the size of the same of type of residues are different, e.g. N-terminal or C-terminal residues in a protein sequence, we will only keep the largest residue as the sample.

Return type

dict

Returns

a dictionary of the sample residues and their nodes

getResiduesData()

Get ResidueData objects for the sample residues found in the molecule.

Return type

list(ResidueData)

Returns

list of ResidueData objects

getMoleculeType()

Classify molecule into large molecule, small solvent, or small ions based on number of heavy atp,s and formal charge.

checkMoleculeRings()

Identify rings in the molecule structure. Create a mapping object of identified rings and their nodes.

findStructAromaticBonds()

Find aromatic bonds in the structure, used to update structure graph edge attributes.

isRingEligible(ring)

Is the ring eligible to be considered as a ring based on its size and coarse graining scale and aromaticity, which depends on mapping method, Martini or DPD.

Parameters

ring (structure.Ring) – the ring

Return type

bool

Returns

True if the ring is eligible, False otherwise

getAugmentedStructGraph()

Generate a networkx.Graph object of the molecule structure. Heavy atoms are used as nodes in the graph. The Graph is augmented using the chemical information of the structure: atomic numbers are added as node attributions and bond orders are used as edge attributes.

getBondsOrderAttr(atom)

Get bond pairs and the bond orders for the given atom to be added to networkx.Graph in dict((int,int): {‘bond_order’: int}) format

Parameters

atom (structure.Structure.atom) – the atom

Return type

dict

Returns

dictionary of bond pairs and their orders

getAtomicNumAttr(atom)

Get atomic number of the atom to be added to networkx.Graph in dict(int:{‘atomic_number’}) format

Return type

dict

Returns

a dictionary of the atom index and its atomic number

getAtomicChargeAttr(atom)

Get formal charge of the atom to be added to networkx.Graph in dict(int:{‘formal_charge’}) format

Return type

dict

Returns

a dictionary of the atom index and its atomic charge

getResidues()

Find residues/polymers/repeating units in the molecule. Create dictionaries of all residues by their name and index and all the graph nodes that belong to them.

Return type

dict

Returns

dictionary of all residues by their name and index and list of their graph nodes

getIsomorphicFragments()

Create an isomorphic mapping for the molecule, to be used to translate the sample residue mapping to all other residues.

Return type

dict

Returns

a dictionary of the isomorphic mapping between the sample residues and all other residues

Raises

MappingError – if the isomorphic mapping cannot be found

getNodeToAtomConversion()

Find mapping between self.st_graph nodes indexes to self.struct atom indexes.

Return type

dict

Returns

a dictionary of the atom index and the graph node index

translateMappingToProps()

Translate the mapping solution to the atom properties on the mapped structure. That is each atom in this molecule will have a property of the bead name and number it belongs to, and its order in the bead. This is used to create the CG structure.

addAtomProps(shared_node_types)

Add the grouping atom properties to the mapped structure, for each atom based on its node type.

getBeadsType()

Bead types for each mapped bead in mapping is defined. This includes bead name, the count number of the bead in the molecule. Beads are compared based on their netwrokx graphs.

getMappedBeadInfo(bead_smarts, bead_charge)

Get the information of a mapped bead as requested.

Parameters

  • bead_smarts (bool) – if True, the smarts for each bead is found.

  • bead_charge (bool) – if True, the total charge of the beads is calculated.

Return type

list(MAPPED_BEAD_INFO)

Returns

List of namedtuples of the beads name, smarts, and charge. If smarts and charge are not requested, they are None.

class schrodinger.application.matsci.automated_cg_mapping.DpdMolecule(molecule_st, scale, struct, mol_idx, species, mol_count, mapping_method='dpd')

Bases: schrodinger.application.matsci.automated_cg_mapping.MoleculeData

A DPD molecule class to hold DPD molecule information.

isRingEligible(ring)

Consider rings that are smaller than the CG scale.

Parameters

ring (structure.Ring) – the ring

Return type

bool

Returns

True if the ring is eligible, False otherwise

class schrodinger.application.matsci.automated_cg_mapping.MartiniMolecule(*args, mapping_method='martini')

Bases: schrodinger.application.matsci.automated_cg_mapping.MoleculeData

A Martini molecule class to hold Martini molecule information.

class RING_INFO(ring_graph, short_end, long_end, centrality)

Bases: tuple

centrality

Alias for field number 3

long_end

Alias for field number 2

ring_graph

Alias for field number 0

short_end

Alias for field number 1

__init__(*args, mapping_method='martini')

Initialize the instance.

processMoleculeInfo()

Overwrite the method to process Martini fragments.

findFragments()

Get ResidueData objects for the sample residues found in the molecule.

Return type

list(ResidueData)

Returns

list of ResidueData objects

classifyMartiniFragments(nodes)

Classify residue nodes into rings with topological symmetry and other fragments.

Parameters

nodes (list) – list of nodes that belong to the residue

Return type

tuple(list, list)

Returns

list of symm ring fragments and list of all otehr fragments

isRingEligible(ring)

Decide if the ring is eligible to be considered as a ring through mapping.

Parameters

ring (structure.Ring) – the ring

Return type

bool

Returns

True if the ring is eligible, False otherwise

class schrodinger.application.matsci.automated_cg_mapping.ResidueData(name, graph, mol_st, mol_rings_mapping, scale, mol_func_groups, exclusive_atoms, node_to_atom_idx, predefined_mapping, mapping_method='dpd')

Bases: schrodinger.application.matsci.automated_cg_mapping.StructData

A class to store data for a residue (repeating unit) in the molecule

__init__(name, graph, mol_st, mol_rings_mapping, scale, mol_func_groups, exclusive_atoms, node_to_atom_idx, predefined_mapping, mapping_method='dpd')

Create an instance.

Parameters

  • name (str) – the name of the residue

  • graph (networkx.Graph) – the residue graph

  • mol_st (structure.Structure) – Molecule structure

  • mol_rings_mapping (Mapping) – The mapping of nodes that belong to a ring

  • scale (int) – Coarse graining scale, number of AA atoms per CG bead

  • mol_func_groups (Mapping) – The mapping of all functional groups in the molecule

  • exclusive_atoms (list(int)) – list of all unsharable atoms in the molecule

  • node_to_atom_idx (dict) – A dictionary of node number to atom index

  • predefined_mapping (Mapping) – The mapping of predefined groups in the molecule

  • mapping_method (str) – Coarse graining method

findGroupsFromMolecule(mol_mapping, graph)

Find groups that are mapped in the molecule and match the residue graph.

Parameters

mol_mapping (Mapping) – The mapping of groups in the molecule

Return type

Mapping

Returns

A mapping of groups in the residue

findFuncGroupsFromMolecule(graph)

Find matches of functional groups and predefined groups in the residue.

getContractedGraphAndMapping()

Contracted graph is generated where functional groups are contracted into a single node, in the contracted graph. Contracted mapping is generated to be able to map back contracted graph to the full graph.

getContractedMapping()

Contract any functional groups to a single node, and record the mapping for the contraction

Return type

Mapping

Returns

The contracted mapping for the contraction

getNodeClusterSets()

Clustering nodes with their neighbors at different scales. This is to account for nodes topological and chemical environment. Starting from a seed node and growing the cluster using multiple scales. The set and the scale that achieves best mapping will be used in the driver.

Return type

list(dict)

Returns

A list of dictionaries, one dict for each scale. Each dict contains the seed node as the key and a list of nodes that are similar to the seed as the value.

getTrialScales(size_contracted_graph)

Get the trial scales for the node clusters.

Parameters

size_contracted_graph (int) – The size of the contracted graph

Return type

list(int)

Returns

The trial scales

growNodeCluster(seed_node, scale, graph, mapping, use_degree=False)

Growing a cluster of nodes starting from a seed node and adding its neighbors until the cluster’s size reaches the input scale.

Here the contracted graph should be used to ensure that functional groups are not split. Contracted mapping is needed to find the actual size of the cluster.

Parameters

  • seed_node (int) – The seed node to start growing the cluster from.

  • scale (int) – The scale of the node cluster.

  • graph (networkx.Graph) – The contracted graph.

  • mapping (Mapping) – The contracted mapping.

  • use_degree (bool) – Whether to use the degree of the nodes to calculate the cluster size

Return type

list(int)

Returns

The node cluster as a list of nodes

getClusterSize(cluster, graph, mapping, use_degree)

Calculating the size of a node cluster. The size is the number of nodes in the cluster. If use_degree is True, the sum of the degrees of the nodes minus 2, if use_degree is True.

Parameters

  • cluster (list(int)) – The cluster.

  • graph (networkx.Graph) – The graph.

  • mapping (Mapping) – The mapping.

  • use_degree (bool) – Whether to use the degree of the nodes

Return type

int

Returns

The size of the cluster

expandNodeClusters(mapping, clusters)

For node clusters calculated the full cluster by expanding functional groups beads to their constituents nodes.

Parameters

  • mapping (Mapping) – The contracted mapping of functional groups to the full structure nodes

  • clusters (dict) – Dictionary of node clusters

Return type

dict

Returns

Dictionary of node clusters with the functional groups expanded

getColoredMapping(node_clusters)

Get a mapping, based on an initial guess from the vertex coloring solution to the dual graph of the input graph of CG functional groups.

Parameters

node_clusters (dict) – dictionary of nodes in the graph and their clusters

Return type

Mapping

Returns

The new mapping

class schrodinger.application.matsci.automated_cg_mapping.MartiniResidueData(*args, mapping_method='martini')

Bases: schrodinger.application.matsci.automated_cg_mapping.ResidueData

A MArtini residue, or a fragment that is not a topologically symmetric ring, e.g. a non symmetrical ring fragment or a non-ring fragment.

__init__(*args, mapping_method='martini')

Initialize the instance.

class schrodinger.application.matsci.automated_cg_mapping.SymmRingFragmentData(ring_info, *args)

Bases: schrodinger.application.matsci.automated_cg_mapping.ResidueData

A Martini fragment that is a topologically symmetric ring.

class RING_NODE_TYPE(index, score, similar_neigh, nodes)

Bases: tuple

index

Alias for field number 0

nodes

Alias for field number 3

score

Alias for field number 1

similar_neigh

Alias for field number 2

__init__(ring_info, *args)

Initialize an instance.

getFragRingMapping()

Get a mapping of the nodes of this fragment that are in a ring. Note we need to use the molecule rings mapping and filter nodes that belong to this fragment, since we don’t have acess to the ring structure. Also note that not all nodes on this fragment are in a ring, they can be on the edge or the dangling ends of the ring fragment.

Return type

Mapping

Returns

The mapping of the nodes of this fragment that are in a ring

classifyRingNodeTypes()

Classify nodes based on their centrality score and topology of the ring. Create RING_NODE_TYPE objects for all nodes of the same centrality score, and track if there are nodes of the same type that are neighbors. Store the information in a list of RING_NODE_TYPE objects.

isSharedNode(node)

Check if the node is shared between two or more rings.

Parameters

node (int) – The node to check

Return type

bool

Returns

Whether the node is shared

getMappingOrderPermulations()

Get order of mapping node types in the ring. And if they should be joined with their same type neighbors or not. This is used to generate all possible permutations of node types in the ring for mapping, to ensure symmetry and topology of the structure is maintained through mapping.

Return type

list(list(NODE_TYPE, bool))

Returns

A list of all possible permutations of node types in the ring

getRingOrderPermutations()

Generate all possible permutations of node types in the ring for mapping, to ensure symmetry and topology of the structure is maintained through mapping. E.g. a permulation list of [[X, Y], [Y, X]] means first we map all nodes of type X (centrality score of X) with all their neighbors and proceed to nodes of type Y. In the second round, we start mapping nodes of type Y and proceed to nodes of type X.

Return type

list(list)

Returns

A list of all possible permutations of nodes in the ring

swapOrder(order, pair1, pair2)

Swap positions of two elements in a list. pair1 and pair2 can be either a single swap (int) or a double swap (tuple).

Parameters

  • order (list) – The list to swap elements in

  • pair1 (int or tuple(int, int)) – The RING_NODE_TYPE.index of the first node type to be swapped with pair2 or a tuple of two indexes to be swapped with each other.

  • pair2 (int or tuple(int, int)) – The RING_NODE_TYPE.index of the first node type to be swapped with pair1 or a tuple of two indexes to be swapped with each other.

Return type

list

Returns

The new list with the elements swapped

class schrodinger.application.matsci.automated_cg_mapping.MappingValidator(fragment, scale)

Bases: object

A validator object used to score CG mappings.

__init__(fragment, scale)

Create an instance. :param fragment: The fragment that is being mapped, can be a whole molecule or a residue. :type fragment: MoleculeData or ResidueData

Parameters

scale (int) – Coarse graining scale, number of AA atoms per CG bead

validateMapping(mapping)

Validate a mapping.

Parameters

mapping (Mapping) – the mapping to be validated

Return type

bool

Returns

whether the mapping is valid

validateNodes(mapping)

validate nodes in the mapping

Parameters

mapping (Mapping) – the mapping to be validated

Return type

bool

Returns

whether the mapping is valid

validateBeads(mapping)

Validate beads in the mapping

Parameters

mapping (Mapping) – the mapping to be validated

Return type

bool

Returns

whether the mapping is valid

validateFuncGroups(mapping)

Validate functional groups are intact

Parameters

mapping (Mapping) – The mapping to be validated

Return type

bool

Returns

whether functional groups are intact

validatePredefinedMapping(mapping)

Validate that pre-defined groups are not expanded.

isScoreImproved(new_score, best_score, new_mapping, best_mapping)

Validate if the score and the new mapping is improved compared to the best mapping from the previous iteration.

Parameters

  • new_score (float) – the new score

  • best_score (float) – the best score from previous iterations

  • new_mapping (Mapping) – the new mapping

  • best_mapping (Mapping) – the best mapping from previous iterations

Return type

bool

Returns

whether the score and mapping have improved

numsOfSharedNodes(mapping, graph)

Find number of shared nodes in the mapping

Parameters

  • mapping (Mapping) – The mapping

  • graph (networkx.Graph) – The graph

Return type

int

Returns

number of shared nodes in the mapping

validateSharesWithNeighbors(node, neighbors, mapping, residue_graph)

Validate number of shared nodes between two beads against node degrees in the graph.

Parameters

  • node (int) – the shared node

  • neighbors (list(int)) – shared nodes neighbors

  • mapping (Mapping) – the mapping to be validated

  • residue_graph (networkx.Graph) – the graph of the residue (or the whole molecule)

Return type

bool

Returns

whether the mapping is valid

validateUnsharedNeighbors(node, neighbors_list, mapping)

Validate whether shared nodes have at least one unshared neighbors in each bead that they belong to.

Parameters

  • node (int) – the shared node

  • neighbors_list (list(int)) – shared nodes neighbors

  • mapping (Mapping) – the mapping to be validated

Return type

bool

Returns

whether the mapping is valid

computeCostFunction(mapping)

Compute the objective function for a given group mapping.

Inspired from T. Bereau and K. Kremer, JCTC 2015. See eq. 1 and the surrounding discussion

Parameters

mapping (Mapping) – CG mapping dict for a section of (or an entire) molecule

Return type

float

Returns

Objective function score

getNumBrokenRings(mapping)

Calculate the number of rings whose CG representation involves more than one group. Attempts to favor CG mappings that center groups on rings (if rings are present)

Parameters

mapping (Mapping) – CG mapping for a section of (or an entire) molecule

Return type

int

Returns

number of broken rings

class schrodinger.application.matsci.automated_cg_mapping.BaseMapper(molecule, scale)

Bases: object

Class for mapping a molecule to a DPD coarse-grained model

__init__(molecule, scale)

Create an instance.

Parameters

  • scale (int) – Coarse graining scale, number of AA atoms per CG bead

  • molecule (MoleculeData) – the molecule to be mapped

map()

Map the AA molecule to a coarse-grained model, depending on the coarse-graining method, e.g. DPD or Martini.

mapResidue(residue)

Map a residue to a DPD coarse-grained model

Parameters

residue (ResidueData) – the residue to be mapped

Return type

tuple(Mapping, dict)

Returns

the best mapping and the best expanded node clusters for the residue

homogenizeBeadSizes(residue, mapping, symmetry_mapping, validator)

Homogenize the bead sizes of a residue in iterations until the score stops improving. The homogenization is done by joining small beads with their neighbors, or swapping nodes from the largest neighbors to the smallest beads.

Parameters

  • residue (ResidueData) – Residue object that holds the residue data

  • mapping (Mapping) – the mapping to be homogenized

  • symmetry_mapping (dict) – a dictionary of classified node clusters

  • validator (MappingValidator) – a validator object to validate the mapping

Return type

Mapping

Returns

Homogenized mapping

moveNodesInIteration(mapping, score, residue, symmetry_mapping, validator, join=True)

Move nodes between beads in iteration in order to homogenize bead sizes. Finds the smallest groups (acceptors) in the current mapping, sorts the set of acceptors according to their symmetry groups, and then for each symmetric subset iterates over symmetric sets and eligible neighboring groups and moves nodes between beads by either joining them or swapping one of their nodes to another bead.

Parameters

  • mapping (Mapping) – the mapping to be homogenized

  • score (float) – the score of the mapping

  • residue (ResidueData) – Residue object that holds the residue data

  • symmetry_mapping (dict) – a dictionary of classified node clusters

  • validator (MappingValidator) – a validator object to validate the mapping

  • join (bool) – whether to join beads or swap nodes between beads

Return type

tuple(Mapping or NoneType, float)

Returns

The best mapping and score for the iteration over acceptor groups. If score is not improved, returned mapping will be None.

getMartiniEdgesInMapping(residue, mapping, cg_graph)

Find groups that are rings in the mapping.

Parameters
Return type

list

Returns

A list of groups indexes that are rings in the mapping

joinBeads(temp_mapping, joiners, joinees, cg_graph, validator, best_score, martini_edges=None)

Join the smallest beads in the mapping with their smallest neighbors.

Parameters

  • temp_mapping (Mapping) – the temporary mapping at this iteration

  • joiners (list) – list of joiner beads

  • joinees (list) – list of joinee beads

  • cg_graph (networkx.Graph) – the coarse-grained graph

  • validator (MappingValidator) – a validator object to validate the mapping

  • best_score (float) – the best score achieved so far

  • martini_edges – list of groups indexes that are rings in the mapping, used when mapping a Martini fragment

Type

list or NoneType

Return type

tuple(Mapping or None, float)

Returns

The new mapping with small beads joined together and the score for this mapping. If score is not improved, returned mapping will be None.

swapBeads(temp_mapping, acceptors, donors, residue, cg_graph, validator, best_score, martini_edges=None)

Performs one iteration of the homogenizeGroupSizes node-swap stage. Finds the smallest groups (acceptors) in the current mapping, sorts the set of acceptors according to their symmetry, and then for each symmetric subset iterates over symmetric sets of the largest neighboring groups (donors) to find the best choice for the node-swap.

Parameters

  • temp_mapping (Mapping) – temporary mapping at the current iteration

  • acceptors (list) – list of acceptor beads

  • donors (list) – list of donor beads

  • residue (ResidueData) – The residue object

  • cg_graph (networkx.Graph) – The coarse-grained graph

  • validator (MappingValidator) – a validator object to validate the mapping

  • best_score (float) – the best score so far

  • martini_edges – list of groups indexes that are rings in the mapping, used when mapping a Martini fragment

Type

list or NoneType

Return type

tuple(Mapping or NoneType, float)

Returns

The best swap mapping and swap score for the iteration over donor groups. If score is not improved, returned mapping will be None.

sharingNodesInSymmRing(swap_pair, acceptors, mapping, residue)

Decide if the swapping pairs will share or donate nodes to each other, for a Martini symmetric ring.

Parameters

  • swap_pair (tuple) – The pair of group ids that will undergo node swap

  • acceptors (list) – List of acceptor beads

  • mapping (Mapping) – The cg mapping at the current iteration

  • residue (ResidueData) – The residue object that is being mapped

Return type

bool or NoneType

Returns

If True, the swapping pairs will share nodes, otherwise nodes are donated. None for fragments that are not Martini rings.

getSwaps(graph)

For a given graph, decompose nodes to list of pairs of beads that will undergo swaps.

Parameters

graph (networkx.Graph) – The subgraph

Return type

list(tuple(int, int))

Returns

List of pairs of beads that will undergo swaps

mapSubgraph(subgraph, new_mapping, is_symm_ring=False)

Use networkx.coloring to cluster nodes in a graph.

Parameters

  • subgraph (networkx.Graph) – The subgraph to be clustered

  • new_mapping (Mapping) – The new mapping

  • is_symm_ring (bool) – Whether the subgraph is a symmetric ring

Return type

Mapping

Returns

The new mapping

processSwapPair(swap, acceptors, mapping)

Process the swapping pair.

Parameters

  • swap (tuple) – The pair of group ids that will undergo node swap

  • acceptors (list) – List of acceptor beads

  • mapping (Mapping) – The cg mapping at the current iteration

Return type

tuple(int, int, set, list)

Returns

The acceptor and donor indexes, the nodes that will undergo swap and the nodes that are already shared between the swapping pairs.

performSwapping(acceptors, swap, mapping, residue, symm_ring_shared)

Carry out the given swap move. If any nodes are already shared between the swapping pairs, the shared node will be donated to the acceptor. Otherwise, a node will be shared between the two.

Parameters

  • acceptors (list(int)) – List of acceptor beads

  • mapping (Mapping) – The cg mapping at the current iteration

  • swap (list) – The pair of group ids that will undergo node swap

  • residue (ResidueData) – The residue object that is being mapped

  • symm_ring_shared – For swaps in symmetric rings, whether the swapping pairs share nodes or donates nodes to each other. None for all other swaps.

Type

bool or NoneType

Return type

Mapping

Returns

The new mapping

fixDisconnectedDonors(acceptor_idx, donor_idx, mapping, residue)

Check if swapping nodes caused the donor group being broken (becoming disconnected). If so, the largest connected component of the resulting donor group becomes the new donor group, and any remaining connected components of the broken donor group become part of the acceptor group.

Parameters

  • acceptor_idx (int) – Group id for the acceptor

  • donor_idx (int) – Group id for the donor

  • mapping (Mapping) – The current mapping

  • residue (ResidueData) – The residue object that is being mapped

Return type

mapping

Returns

The new mapping

shareNodes(accept_idx, donor_idx, swap, mapping, residue)

Any nodes that are to be swapped and are not already shared, become shared here, unless that are marked as ‘exclusive’ (i.e. not to be shared). If a node is exclusive, it is instead completely donated from the donor to the acceptor group. In this case, must also check if the node is part of a functional group, in which case the entire functional group is donated from the donor to the acceptor.

Parameters

  • accept_idx (int) – The acceptor bead index

  • donor_idx (int) – The donor bead index

  • swap (list(int)) – The list of nodes that are already shared between donor and acceptor

  • mapping (Mapping) – The cg mapping at the current iteration

  • residue (ResidueData) – The residue object that is being mapped

Return type

Mapping

Returns

The new mapping

donateNodes(accept_idx, donor_idx, shared_nodes, mapping)

Any nodes that are already shared between the acceptor and the donor groups, become exclusive to the acceptor group. If a given shared node belongs to a functional group, the entire functional group must be transferred from the donor to the acceptor.

Parameters

  • accept_idx (int) – The acceptor bead index

  • donor_idx (int) – The donor bead index

  • shared_nodes (list(int)) – The list of nodes that are already shared between donor and acceptor

  • mapping (Mapping) – The cg mapping at the current iteration

Return type

Mapping

Returns

The new mapping

moveNodesBetweenBeads(bead, acceptor, donor, mapping)

Move all nodes in the bead (functional group) from donor to the acceptor.

Parameters

  • bead (int) – the functional group bead

  • acceptor (int) – The acceptor bead index

  • donor (int) – The donor bead index

  • mapping (Mapping) – The mapping that should be modified

Return type

Mapping

Returns

The new mapping

categorizeSymmetryMappings(mapping, symmetry_mapping)

Categorize CG beads into same groups based on the symmetry mapping. This is to ensure that symmetrical beads are all homogenized the same.

Parameters

  • mapping (Mapping) – bead mapping

  • symmetry_mapping (Mapping) – Mapping of the category ids of individual nodes

Return type

dict

Returns

dictionary of beads and the sorted symmetry ids of the nodes that are grouped in that bead.

groupSymmetricBeads(beads, cg_symmetry_dict)

Sort and group beads based on their symmetry ids.

Parameters

  • beads (list) – beads to be sorted

  • cg_symmetry_dict (dict) – dictionary of beads and the symmetry ids of the nodes that are grouped in that bead.

Return type

dict

Returns

dictionary of symmetry ids and the beads that belong to the symmetry group

findGroupsInSet(residue, mapping, beads=None, smallest=True)

Find group of beads in the mapping based on their size.

Parameters

  • residue (ResidueData) – the residue to be mapped

  • mapping (Mapping) – bead mapping

  • beads (list) – list of beads to search within. If None, search all beads

  • smallest (bool) – if True, find the smallest beads, else find the largest

Return type

tuple(list, int)

Returns

the eligible beads and their size

expandResidualMapping(residue_mapping, residue_node_clusters=None)

Expand the mapping of the sample residue to all other residues it represents.

Parameters

  • residue_mapping (dict) – Mapping for each sample residue

  • residue_node_clusters (dict or None) – Best node cluster found for each residue

Return type

Mapping

Returns

Mapping for all residues

class schrodinger.application.matsci.automated_cg_mapping.DpdMapper(molecule, scale)

Bases: schrodinger.application.matsci.automated_cg_mapping.BaseMapper

Mapper for DPD coarse-grained model

map()

Map the molecule to a DPD coarse-grained model

getSwaps(graph)

For a given graph, decompose nodes to list of pairs of beads that will undergo swaps.

Parameters

graph (networkx.Graph) – The subgraph

Return type

list(tuple(int, int))

Returns

List of pairs of beads that will undergo swaps

class schrodinger.application.matsci.automated_cg_mapping.MartiniMapper(molecule, scale=4)

Bases: schrodinger.application.matsci.automated_cg_mapping.BaseMapper

__init__(molecule, scale=4)

Initialize the MartiniMapper. The scale is set to MARTINI_SCALE for a Martini model.

map()

Map the molecule to a Martini coarse-grained model

homogenizeRingBeadSizes(residue, mapping, symmetry_mapping, validator)

Homogenize the bead sizes of a residue in iterations until the score stops improving. The homogenization is done by swapping nodes from the largest neighbors to the smallest beads.

Parameters

  • residue (ResidueData) – Residue object that holds the residue data

  • mapping (Mapping) – the mapping to be homogenized

  • symmetry_mapping (dict) – a dictionary of classified node clusters

  • validator (MappingValidator) – a validator object to validate the mapping

Return type

Mapping

Returns

Homogenized Martini Rings mapping

mapSymmRingFragment(fragment)

Map Martini fragments with a symmetrical ring. Ring nodes are classified based on their centrality score as a measure of the ring topology. To map, we generate different permutations of grouping nodes types and group nodes in the ring. The best mapping is selected by swaping nodes between beads until best score is achieved.

Parameters

fragment (SymmRingFragmentData) – fragment object that holds the fragment data

Return type

Mapping

Returns

The best mapping

getMappingFromOrder(fragment, order)

Map nodes into beads based on the order of nodes types given in order list.

Parameters

  • fragment (SymmRingFragmentData) – fragment object that holds the fragment data

  • order (list) – The order of the grouping

Return type

Mapping

Returns

The mapping

getSwaps(graph)

For a given graph, all nodes that are connected by an edge will be added as a swaping pair.

Parameters

graph (networkx.Graph) – The subgraph

Return type

list(tuple(int, int))

Returns

list of pairs of beads that will undergo swaps

class schrodinger.application.matsci.automated_cg_mapping.MappingMethod

Bases: object

Class to decide the class bundle corresponding to a give CG mapping method.

METHOD_BUNDLE

alias of schrodinger.application.matsci.automated_cg_mapping.MethodBundle

METHODS = {'dpd': MethodBundle(mol_data=<class 'schrodinger.application.matsci.automated_cg_mapping.DpdMolecule'>, res_data=<class 'schrodinger.application.matsci.automated_cg_mapping.ResidueData'>, mapper=<class 'schrodinger.application.matsci.automated_cg_mapping.DpdMapper'>, scale=None), 'martini': MethodBundle(mol_data=<class 'schrodinger.application.matsci.automated_cg_mapping.MartiniMolecule'>, res_data=<class 'schrodinger.application.matsci.automated_cg_mapping.MartiniResidueData'>, mapper=<class 'schrodinger.application.matsci.automated_cg_mapping.MartiniMapper'>, scale=4)}
static molecule(method)

Get the molecule classe based on the method.

Parameters

method (str) – The method name

Return type

DPDMoleculs or MartiniMolecule

Returns

The molecule class

static residue(method)

Get the residue class based on the method.

Parameters

method (str) – The method name

Return type

ResidueData or MartiniResidueData

Returns

The residue class

static mapper(method)

Get the mapper class based on the method.

Parameters

method (str) – The method name

Return type

DpdMapper or MartiniMapper

Returns

The mapper class

static scale(method, inp_scale)

Get the scale based on the method.

Parameters

  • method (str) – The method name

  • inp_scale (int) – The input scale

Return type

int

Returns

The scale

class schrodinger.application.matsci.automated_cg_mapping.AutomatedCGMapping(struct, scale=4, bead_smarts=False, bead_charge=False, predefined_patterns=None, method='dpd')

Bases: object

Main class to map the atomistic structure to coarse-grained

__init__(struct, scale=4, bead_smarts=False, bead_charge=False, predefined_patterns=None, method='dpd')

Create an instance.

Parameters

  • struct (structure.Structure) – All-atom structure to be mapped to CG

  • scale (int) – Coarse graining scale, number of AA atoms per CG bead.

  • bead_smarts (bool) – If True, will get SMARTS for found beads.

  • bead_charge (bool) – If True, will get charges for found beads.

  • predefined_patterns (dict) – dictionary of names and pre-defined SMARTS patterns that need to be mapped in one bead

  • method (str) – The method to be used for mapping

map()

Run the workflow.