sccross.models.sccross.SCCROSSModel¶

class sccross.models.sccross.SCCROSSModel(adatas, latent_dim=50, h_depth=2, h_dim=256, dropout=0.2, shared_batches=False, random_seed=0)[source]¶

Bases: sccross.models.utils.Model

Methods

`adopt_pretrained_model`	Adopt buffers and parameters from a pretrained model
`compile`	Prepare model for training
`encode_data`	Compute data (cell) embedding
`fit`	Fit model on given datasets
`freeze_cells`	Freeze cell embeddings
`generate_augment`	Generate augmented single cell omic data
`generate_cross`	Cross generation
`generate_multiSim`	Generate multi-omics simulation data
`get_losses`	Compute loss function values
`perturbation_difGenes`	Obtain differential genes generated by perturbation
`unfreeze_cells`	Unfreeze cell embeddings

Attributes

`ALIGN_BURNIN_PRG`
`GRAPH_BATCHES`
`MAX_EPOCHS_PRG`
`PATIENCE_PRG`
`REDUCE_LR_PATIENCE_PRG`
`logger`