Model#

The DRVI (Disentangled Representation Variational Inference) model is a model designed for single-cell omics data analysis. It provides disentangled latent representations that separate individual biological processes, enabling better interpretation and downstream analysis.

Overview#

DRVI extends the standard variational autoencoder architecture with specialized decoder architecture. The model learns disentangled representations and separates different sources of variation in the data, such as:

Biological factors: Cell types, developmental processes, perturbation responses, signaling pathways
Technical factors: Background expressions, technical stress responses

Core Components#

DRVI model#

This is the main model class that can be used to define, train, and evaluate the model on an anndata. DRVI passes any extra argument to DRVIModule in initialization. Accordingly, we suggest to check its documentation (below) for additional configurations.

model.DRVI

DRVI model based on scvi-tools framework for disentangled representation learning.

DRVIModule#

This is the pytorch neural network module and contains DRVI logic.

model.DRVIModule

DRVI (Disentangled Representation Variational Inference) pytorch module.

Usage Example#

import anndata as ad
from drvi.model import DRVI

# Load your data
adata = ad.read_h5ad("your_data.h5ad")

# Setup anndata
DRVI.setup_anndata(
    adata,
    layer="counts",
    categorical_covariate_keys=["batch"],
    is_count_data=True,
)

# Initialize the model
model = DRVI(
    adata,
    n_latent=64,
    encoder_dims=[128, 128],
    decoder_dims=[128, 128],
)

# Train the model
model.train(
    max_epochs=400,
    early_stopping=False,
)

# Get disentangled representations
latent = model.get_latent_representation()

# Please check tutorials for more details and downstream steps

Model

Contents