Subgraph Expansion Workflow¶

pybel_tools.mutation.expansion.get_peripheral_successor_edges(graph, subgraph)[source]¶

Gets the set of possible successor edges peripheral to the subgraph. The source nodes in this iterable are all inside the subgraph, while the targets are outside.

Parameters:	graph (pybel.BELGraph) – A BEL graph subgraph – An iterator of BEL nodes
Returns:	An iterable of possible successor edges (4-tuples of node, node, key, data)
Return type:	iter[tuple]

pybel_tools.mutation.expansion.get_peripheral_predecessor_edges(graph, subgraph)[source]¶

Gets the set of possible predecessor edges peripheral to the subgraph. The target nodes in this iterable are all inside the subgraph, while the sources are outside.

Parameters:	graph (pybel.BELGraph) – A BEL graph subgraph – An iterator of BEL nodes
Returns:	An iterable on possible predecessor edges (4-tuples of node, node, key, data)
Return type:	iter[tuple]

pybel_tools.mutation.expansion.count_sources(edge_iter)[source]¶

Counts the source nodes in an edge iterator with keys and data

Parameters:	edge_iter (iter[tuple]) – An iterable on possible predecessor edges (4-tuples of node, node, key, data)
Returns:	A counter of source nodes in the iterable
Return type:	collections.Counter

pybel_tools.mutation.expansion.count_targets(edge_iter)[source]¶

Counts the target nodes in an edge iterator with keys and data

Parameters:	edge_iter (iter[tuple]) – An iterable on possible predecessor edges (4-tuples of node, node, key, data)
Returns:	A counter of target nodes in the iterable
Return type:	collections.Counter

pybel_tools.mutation.expansion.count_possible_successors(graph, subgraph)[source]¶

Parameters:	graph (pybel.BELGraph) – A BEL graph subgraph – An iterator of BEL nodes
Returns:	A counter of possible successor nodes
Return type:	collections.Counter

pybel_tools.mutation.expansion.count_possible_predecessors(graph, subgraph)[source]¶

Parameters:	graph (pybel.BELGraph) – A BEL graph subgraph – An iterator of BEL nodes
Returns:	A counter of possible predecessor nodes
Return type:	collections.Counter

pybel_tools.mutation.expansion.get_subgraph_edges(graph, annotation, value, source_filter=None, target_filter=None)[source]¶

Gets all edges from a given subgraph whose source and target nodes pass all of the given filters

Parameters:	graph (pybel.BELGraph) – A BEL graph annotation (str) – The annotation to search value (str) – The annotation value to search by source_filter – Optional filter for source nodes (graph, node) -> bool target_filter – Optional filter for target nodes (graph, node) -> bool
Returns:	An iterable of (source node, target node, key, data) for all edges that match the annotation/value and node filters
Return type:	iter[tuple]

pybel_tools.mutation.expansion.get_subgraph_peripheral_nodes(graph, subgraph, node_filters=None, edge_filters=None)[source]¶

Gets a summary dictionary of all peripheral nodes to a given subgraph

Parameters:	graph (pybel.BELGraph) – A BEL graph subgraph (iter[tuple]) – A set of nodes node_filters (lambda) – Optional. A list of node filter predicates with the interface (graph, node) -> bool. See `pybel_tools.filters.node_filters` for more information edge_filters (lambda) – Optional. A list of edge filter predicates with the interface (graph, node, node, key, data) -> bool. See `pybel_tools.filters.edge_filters` for more information
Returns:	A dictionary of {external node: {‘successor’: {internal node: list of (key, dict)}, ‘predecessor’: {internal node: list of (key, dict)}}}
Return type:	dict

For example, it might be useful to quantify the number of predecessors and successors

>>> import pybel_tools as pbt
>>> sgn = 'Blood vessel dilation subgraph'
>>> sg = pbt.selection.get_subgraph_by_annotation_value(graph, annotation='Subgraph', value=sgn)
>>> p = pbt.mutation.get_subgraph_peripheral_nodes(graph, sg, node_filters=pbt.filters.exclude_pathology_filter)
>>> for node in sorted(p, key=lambda n: len(set(p[n]['successor']) | set(p[n]['predecessor'])), reverse=True):
>>>     if 1 == len(p[sgn][node]['successor']) or 1 == len(p[sgn][node]['predecessor']):
>>>         continue
>>>     print(node,
>>>           len(p[node]['successor']),
>>>           len(p[node]['predecessor']),
>>>           len(set(p[node]['successor']) | set(p[node]['predecessor'])))

pybel_tools.mutation.expansion.expand_periphery(universe, graph, node_filters=None, edge_filters=None, threshold=2)[source]¶

Iterates over all possible edges, peripheral to a given subgraph, that could be added from the given graph. Edges could be added if they go to nodes that are involved in relationships that occur with more than the threshold (default 2) number of nodes in the subgraph.

Parameters:

universe (pybel.BELGraph) – The universe of BEL knowledge
graph (pybel.BELGraph) – The (sub)graph to expand
node_filters (lambda) – Optional. A list of node filter predicates with the interface (graph, node) -> bool. See pybel_tools.filters.node_filters for more information
edge_filters (lambda) – Optional. A list of edge filter predicates with the interface (graph, node, node, key, data) -> bool. See pybel_tools.filters.edge_filters for more information
threshold – Minimum frequency of betweenness occurrence to add a gap node

A reasonable edge filter to use is pybel_tools.filters.keep_causal_edges() because this function can allow for huge expansions if there happen to be hub nodes.

pybel_tools.mutation.expansion.enrich_grouping(universe, graph, function, relation)[source]¶

Adds all of the grouped elements. See enrich_complexes(), enrich_composites(), and enrich_reactions()

Parameters:	universe (pybel.BELGraph) – A BEL graph representing the universe of all knowledge graph (pybel.BELGraph) – The target BEL graph to enrich function (str) – The function by which the subject of each triple is filtered relation (str) – The relationship by which the predicate of each triple is filtered

pybel_tools.mutation.expansion.enrich_complexes(universe, graph)[source]¶

Add all of the members of the complex abundances to the graph.

Parameters:	universe (pybel.BELGraph) – A BEL graph representing the universe of all knowledge graph (pybel.BELGraph) – The target BEL graph to enrich

pybel_tools.mutation.expansion.enrich_composites(universe, graph)[source]¶

Adds all of the members of the composite abundances to the graph.

Parameters:	universe (pybel.BELGraph) – A BEL graph representing the universe of all knowledge graph (pybel.BELGraph) – The target BEL graph to enrich

pybel_tools.mutation.expansion.enrich_reactions(universe, graph)[source]¶

Adds all of the reactants and products of reactions to the graph.

Parameters:	universe (pybel.BELGraph) – A BEL graph representing the universe of all knowledge graph (pybel.BELGraph) – The target BEL graph to enrich

pybel_tools.mutation.expansion.enrich_variants(universe, graph, func=None)[source]¶

Add the reference nodes for all variants of the given function.

Parameters:	universe (pybel.BELGraph) – A BEL graph representing the universe of all knowledge graph (pybel.BELGraph) – The target BEL graph to enrich or iter[str] func (str) – The function by which the subject of each triple is filtered. Defaults to the set of protein, rna, mirna, and gene.

pybel_tools.mutation.expansion.enrich_unqualified(universe, graph)[source]¶

Enrich the subgraph with the unqualified edges from the graph.

Parameters:	universe (pybel.BELGraph) – A BEL graph representing the universe of all knowledge graph (pybel.BELGraph) – The target BEL graph to enrich

The reason you might want to do this is you induce a subgraph from the original graph based on an annotation filter, but the unqualified edges that don’t have annotations that most likely connect elements within your graph are not included.