mosaicmpi.network.Network

class mosaicmpi.network.Network(integration: Integration, subset_nodes: Iterable[str] | None = None, communities: Dict[str, Collection[str]] | None = None)

Create a program network from an Integration object.

Parameters:

• integration (Integration) – Integration of multiple datasets.

• subset_nodes (Iterable[str], optional) – Create an SNS from a subset of the larger program graph, defaults to None

• communities (Dict[str, Collection[str]], optional) – Use pre-defined communities, defaults to None

Attributes

n_communities	Get the number of communities in the Network.
node_to_community
ordered_community_names	Get community names, ordered numerically after separating clusters and subclusters.
programs_in_graph	Get the nodes in the program graph as (dataset, k, program) tuples.

Methods

add_community_weights_to_graph([...])	Add attributes to the program graph for generating the community-weighted network.
community_search([algorithm, resolution, k, ...])	Identifies communities from the program graph.
compute_community_network_layout([...])	_summary_
compute_layout([algorithm, base_weight, ...])	Compute the network layout using a specified algorithm.
consensus([method, min_k])	Generate a 'consensus' program for each community and dataset by taking the median of all constituent programs, separately for each dataset.
count_intracommunity_edges()	Counts edges within each community that are within and between datasets.
create_community_network()	Creates community network after community search.
create_program_network()	Creates a program graph based on pairwise correlation thresholds and selected ranks.
from_pkl(filename)	Read an Network object from a file.
get_community_category_overrepresentation(layer)	_summary_
get_community_metadata_correlation(layer[, ...])	Calculate Pearson correlation of program usage to numerical metadata across samples/observations.
get_community_usage([subset_datasets, ...])	Get median usage of each community of programs for each samples.
get_lowest_rank_programs([min_k])	Identify the programS that are the lowest rank for each dataset.
get_max_corr_communities()	Create a matrix with community and dataset on each axis.
get_node_table()	Get node counts before and after various node and edge filters.
get_representative_program_ids([...])	Select programs based on correlation with the median of all programs in each community
get_representative_programs([correlation_axis])	Select programs based on correlation with the median of all programs in each community
get_sample_entropy([subset_datasets])	Get shannon diversity of Community Usage for each sample.
get_selected_rank_program_ids(k)	Select programs based on rank.
get_vectorized_community_sort_key(...)	Return a vector of sort_indicesGet community names, ordered numerically after separating clusters and subclusters.
most_correlated_edge_between_datasets(ds1, ds2)	Identifies the most correlated edge between two datasets within each community.
prune_communities([min_nodes, min_datasets, ...])	Prune communities based on one or more filters.
read_communities_from_toml(toml_file)
to_pkl(filename)	Persists the SNS object using python's pickle format with optional gzip compression :param filename: path to .pkl or .pkl.gz file :type filename: str
transfer_labels([source, dest, layer, ...])	Transfer sample categories between datasets using usage of representative programs as a proxy.
write_communities_toml(filename)	Write communities to TOML file.
write_community_network_graphml(filename)	Output the community network in graphml format.
write_program_network_graphml(filename)	Output the program network in graphml format.