scglue.models.scglue.SCGLUEModel
- class scglue.models.scglue.SCGLUEModel(adatas, vertices, latent_dim=50, h_depth=2, h_dim=256, dropout=0.2, shared_batches=False, random_seed=0)[source]
Bases:
Model
GLUE model for single-cell multi-omics data integration
- Parameters
adatas (
typing.Mapping
[str
,AnnData
]) – Datasets (indexed by modality name)vertices (
typing.List
[str
]) – Guidance graph vertices (must cover feature names in all modalities)latent_dim (
int
) – Latent dimensionalityh_depth (
int
) – Hidden layer depth for encoder and discriminatorh_dim (
int
) – Hidden layer dimensionality for encoder and discriminatordropout (
float
) – Dropout rateshared_batches (
bool
) – Whether the same batches are shared across modalitiesrandom_seed (
int
) – Random seed
Methods
Adopt buffers and parameters from a pretrained model
Prepare model for training
Decode data
Compute data (cell) embedding
Compute graph (feature) embedding
Fit model on given datasets
Freeze cell embeddings
Compute loss function values
Unfreeze cell embeddings
Upgrade the model if generated by older versions
Attributes
ALIGN_BURNIN_PRG
GRAPH_BATCHES
MAX_EPOCHS_PRG
PATIENCE_PRG
REDUCE_LR_PATIENCE_PRG
logger