network_analysis

Function

crisgi_obj.network_analysis(
    target_group,
    layer="log1p",
    method="prod",
    test_type="TER",
    interactions=None,
    unit_header="subject",
    out_dir=None,
    n_neighbors=10,
    strategy="bottom_up",
)

Performs network analysis on the dataset using specified parameters and interaction sets. This method extracts and analyzes interaction features from the data, supporting various strategies and test types for flexible network exploration.

Parameters

Name	Type	Description
target_group	str	The group identifier for which the network analysis is performed.
layer	str, optional	The data layer to use for analysis (default: 'log1p').
method	str, optional	The method for interaction calculation (default: 'prod').
test_type	str, optional	The statistical test type to use (default: 'TER').
interactions	list, optional	List of interaction features to include; if None, uses default from edata.uns.
unit_header	str, optional	The header indicating the unit of analysis (default: 'subject').
out_dir	str, optional	Output directory for results; if None, results are not saved to disk.
n_neighbors	int, optional	Number of neighbors to consider in the analysis (default: 10).
strategy	str, optional	Strategy for network construction (default: 'bottom_up').

Return type

None

Returns

This function does not return a value. It performs network analysis and may print or save intermediate results depending on the parameters.

Attributes Set

• May access or update self.edata, self.adata, and related attributes.
• Uses or sets keys in edata.uns.

Example

# Assume `crisgi` is an instance of the CRISGI class

# Perform network analysis for group 'A' with default settings
crisgi.network_analysis(target_group='A')

# Perform network analysis with custom interactions and output directory
custom_interactions = ['gene1', 'gene2', 'gene3']
crisgi.network_analysis(
    target_group='B',
    interactions=custom_interactions,
    out_dir='/path/to/output',
    n_neighbors=15,
    strategy='top_down'
)