omicverse.single.pyCEFCON

omicverse.single.pyCEFCON#

class omicverse.single.pyCEFCON(input_expData, input_priorNet, input_genesDE=None, additional_edges_pct=0.01, cuda=0, seed=2023, hidden_dim=128, output_dim=64, heads=4, attention='COS', miu=0.5, epochs=350, repeats=5, edge_threshold_param=8, remove_self_loops=False, topK_drivers=100, solver='GUROBI')[source]#

CEFCON workflow wrapper for driver-regulator discovery.

Parameters:

  • input_expData (str or AnnData or pd.DataFrame) – Input expression data with optional lineage metadata.

  • input_priorNet (str or pd.DataFrame) – Prior interaction network.

  • input_genesDE (str or pd.DataFrame, optional) – Differential-expression score table.

  • repeats (int, default=5) – Number of model repeats.

  • solver ({'GUROBI', 'SCIP'}, default='GUROBI') – ILP solver used for driver-regulator selection.

__init__(input_expData, input_priorNet, input_genesDE=None, additional_edges_pct=0.01, cuda=0, seed=2023, hidden_dim=128, output_dim=64, heads=4, attention='COS', miu=0.5, epochs=350, repeats=5, edge_threshold_param=8, remove_self_loops=False, topK_drivers=100, solver='GUROBI')[source]#
Parameters:

  • pd.DataFrame) (input_genesDE (str or)

  • pd.DataFrame)

  • pd.DataFrame)

  • (float (miu)

  • optional) (Solver ('GUROBI', 'SCIP') for solving the integer linear programming problems (for identifying drive regulators) (default: 'GUROBI'))

  • (int (topK_drivers)

  • optional)

  • (int

  • optional)

  • (int

  • optional)

  • (int

  • optional)

  • (int

  • optional)

  • (str (solver)

  • optional)

  • (float

  • optional)

  • (int

  • optional)

  • (int

  • optional)

  • (int

  • optional)

  • (bool (remove_self_loops)

  • optional)

  • (int

  • optional)

  • (str

  • optional)

Return type:

None

Methods

__init__(input_expData, input_priorNet[, ...])

predicted_RGM()

predicted_driver_regulators()

preprocess()

train()

Contents