scvi.external.TorchMRVI

scvi.external.TorchMRVI#

class scvi.external.TorchMRVI(adata=None, registry=None, **model_kwargs)[source]#

Multi-resolution Variational Inference (MrVI) [Boyeau et al., 2025].

Parameters:

adata (AnnData | None (default: None)) – AnnData object that has been registered via setup_anndata().
n_latent – Dimensionality of the latent space for z.
n_latent_u – Dimensionality of the latent space for u.
encoder_n_hidden – Number of nodes per hidden layer in the encoder.
encoder_n_layers – Number of hidden layers in the encoder.
z_u_prior – Whether to use a prior for z_u.
z_u_prior_scale – Scale of the prior for the difference between z and u.
u_prior_scale – Scale of the prior for u.
u_prior_mixture – Whether to use a mixture model for the u prior.
u_prior_mixture_k – Number of components in the mixture model for the u prior.
learn_z_u_prior_scale – Whether to learn the scale of the z and u difference prior during training.
laplace_scale – Scale parameter for the Laplace distribution in the decoder.
scale_observations – Whether to scale loss by the number of observations per sample.
px_kwargs – Keyword args for DecoderZXAttention.
qz_kwargs – Keyword args for EncoderUZ.
qu_kwargs – Keyword args for EncoderXU.

Notes

This implementation of MRVI is in PyTorch. This will become the default version in v1.5 for MRVI.

See further usage examples in the following tutorial:

/tutorials/notebooks/scrna/MrVI_tutorial_torch

See the user guide for this model:

/user_guide/models/mrvi_jax

Attributes table#

`adata`	Data attached to model instance.
`adata_manager`	Manager instance associated with self.adata.
`device`	The current device that the module's params are on.
`get_normalized_function_name`	What the get normalized functions name is
`history`	Returns computed metrics during training.
`is_trained`	Whether the model has been trained.
`minified_data_type`	The type of minified data associated with this model, if applicable.
`registry`	Data attached to model instance.
`run_id`	Returns the run id of the model.
`run_name`	Returns the run name of the model.
`summary_string`	Summary string of the model.
`test_indices`	Observations that are in test set.
`train_indices`	Observations that are in train set.
`validation_indices`	Observations that are in validation set.

Methods table#

`compute_local_statistics`(reductions[, ...])	Compute local statistics from counterfactual sample representations.
`convert_legacy_save`(dir_path, output_dir_path)	Converts a legacy saved model (<v0.15.0) to the updated save format.
`data_registry`(registry_key)	Returns the object in AnnData associated with the key in the data registry.
`deregister_manager`([adata])	Deregisters the `AnnDataManager` instance associated with adata.
`differential_abundance`([adata, ...])	Compute the differential abundance between samples.
`differential_expression`([adata, ...])	Compute cell-specific multivariate differential expression.
`get_aggregated_posterior`([adata, sample, ...])	Computes the aggregated posterior over the `u` latent representations.
`get_anndata_manager`(adata[, required])	Retrieves the `AnnDataManager` for a given AnnData object.
`get_batch_representation`([adata, indices, ...])	Get the batch representation for a given set of indices.
`get_elbo`([adata, indices, batch_size, ...])	Compute the evidence lower bound (ELBO) on the data.
`get_feature_correlation_matrix`([adata, ...])	Generate gene-gene correlation matrix using scvi uncertainty and expression.
`get_from_registry`(adata, registry_key)	Returns the object in AnnData associated with the key in the data registry.
`get_importance_weights`(adata, indices, qz, ...)	Computes importance weights for the given samples.
`get_latent_library_size`([adata, indices, ...])	Returns the latent library size for each cell.
`get_latent_representation`([adata, indices, ...])	Compute the latent representation of the data.
`get_likelihood_parameters`([adata, indices, ...])	Estimates for the parameters of the likelihood \(p(x \mid z)\).
`get_local_sample_distances`([adata, ...])	Compute local sample distances.
`get_local_sample_representation`([adata, ...])	Compute the local sample representation of the cells in the `adata` object.
`get_marginal_ll`([adata, indices, ...])	Compute the marginal log-likehood of the data.
`get_normalized_expression`([adata, indices, ...])	Returns the normalized (decoded) gene expression.
`get_outlier_cell_sample_pairs`([adata, ...])	Compute admissibility scores for cell-sample pairs.
`get_reconstruction_error`([adata, indices, ...])	Compute the reconstruction error on the data.
`get_setup_arg`(setup_arg)	Returns the string provided to setup of a specific setup_arg.
`get_state_registry`(registry_key)	Returns the state registry for the AnnDataField registered with this instance.
`get_var_names`([legacy_mudata_format])	Variable names of input data.
`load`(dir_path[, adata, accelerator, device, ...])	Instantiate a model from the saved output.
`load_query_data`([adata, reference_model, ...])	Online update of a reference model with scArches algorithm [Lotfollahi et al., 2021].
`load_registry`(dir_path[, prefix])	Return the full registry saved with the model.
`minify_adata`([minified_data_type, ...])	Minify the model's `adata`.
`posterior_predictive_sample`([adata, ...])	Generate predictive samples from the posterior predictive distribution.
`prepare_query_anndata`(adata, reference_model)	Prepare data for query integration.
`prepare_query_mudata`(mdata, reference_model)	Prepare multimodal dataset for query integration.
`register_manager`(adata_manager)	Registers an `AnnDataManager` instance with this model class.
`save`(dir_path[, prefix, overwrite, ...])	Save the state of the model.
`setup_anndata`(adata[, layer, sample_key, ...])	Sets up the `AnnData` object for this model.
`to_device`(device)	Move the model to the device.
`train`([max_epochs, accelerator, devices, ...])	Train the model.
`transfer_fields`(adata, **kwargs)	Transfer fields from a model to an AnnData object.
`update_sample_info`(adata)	Initialize/update metadata in the case where additional covariates are added.
`update_setup_method_args`(setup_method_args)	Update setup method args.
`view_anndata_setup`([adata, ...])	Print summary of the setup for the initial AnnData or a given AnnData object.
`view_registry`([hide_state_registries])	Prints summary of the registry.
`view_setup_args`(dir_path[, prefix])	Print args used to setup a saved model.
`view_setup_method_args`()	Prints setup kwargs used to produce a given registry.

Attributes#

TorchMRVI.adata[source]#: Data attached to model instance.

TorchMRVI.adata_manager[source]#: Manager instance associated with self.adata.

TorchMRVI.device[source]#: The current device that the module’s params are on.

TorchMRVI.get_normalized_function_name[source]#: What the get normalized functions name is

TorchMRVI.history[source]#: Returns computed metrics during training.

TorchMRVI.is_trained[source]#: Whether the model has been trained.

TorchMRVI.minified_data_type[source]#: The type of minified data associated with this model, if applicable.

TorchMRVI.registry[source]#: Data attached to model instance.

TorchMRVI.run_id[source]#: Returns the run id of the model. Used in MLFlow

TorchMRVI.run_name[source]#: Returns the run name of the model. Used in MLFlow

TorchMRVI.summary_string[source]#: Summary string of the model.

TorchMRVI.test_indices[source]#: Observations that are in test set.

TorchMRVI.train_indices[source]#: Observations that are in train set.

TorchMRVI.validation_indices[source]#: Observations that are in validation set.

Methods#

TorchMRVI.compute_local_statistics(reductions, adata=None, indices=None, batch_size=None, use_vmap='auto', norm='l2', mc_samples=10)[source]#

Compute local statistics from counterfactual sample representations.

Local statistics are reductions over either the local counterfactual latent representations or the resulting local sample distance matrices. For a large number of cells and/or samples, this method can avoid scalability issues by grouping over cell-level covariates.

Parameters:

reductions (list[MRVIReduction]) – List of reductions to compute over local counterfactual sample representations.
adata (AnnData | None (default: None)) – AnnData object to use.
indices (Union[_Buffer, _SupportsArray[dtype[Any]], _NestedSequence[_SupportsArray[dtype[Any]]], complex, bytes, str, _NestedSequence[complex | bytes | str], None] (default: None)) – Indices of cells to use.
batch_size (int | None (default: None)) – Batch size to use for computing the local statistics.
use_vmap (Literal['auto', True, False] (default: 'auto')) – Whether to use vmap to compute the local statistics. If “auto”, vmap will be used if the number of samples is less than 500.
norm (str (default: 'l2')) – Norm to use for computing the distances.
mc_samples (int (default: 10)) – Number of Monte Carlo samples to use for computing the local statistics. Only applies if using sampled representations.

Return type:

Dataset

classmethod TorchMRVI.convert_legacy_save(dir_path, output_dir_path, overwrite=False, prefix=None, **save_kwargs)[source]#

Converts a legacy saved model (<v0.15.0) to the updated save format.

Parameters:

dir_path (str) – Path to the directory where the legacy model is saved.
output_dir_path (str) – Path to save converted save files.
overwrite (bool (default: False)) – Overwrite existing data or not. If False and directory already exists at output_dir_path, an error will be raised.
prefix (str | None (default: None)) – Prefix of saved file names.
**save_kwargs – Keyword arguments passed into save().

Return type:

None

TorchMRVI.data_registry(registry_key)[source]#

Returns the object in AnnData associated with the key in the data registry.

Parameters:: registry_key (str) – key of an object to get from self.data_registry
Return type:: ndarray | DataFrame
Returns:: The requested data.

TorchMRVI.deregister_manager(adata=None)[source]#

Deregisters the AnnDataManager instance associated with adata.

If adata is None, deregisters all AnnDataManager instances in both the class and instance-specific manager stores, except for the one associated with this model instance.

TorchMRVI.differential_abundance(adata=None, sample_cov_keys=None, sample_subset=None, compute_log_enrichment=False, omit_original_sample=True, batch_size=128)[source]#

Compute the differential abundance between samples.

Computes the logarithm of the ratio of the probabilities of each sample conditioned on the estimated aggregate posterior distribution of each cell.

Parameters:

adata (AnnData | None (default: None)) – The data object to compute the differential abundance for. If not given, the data object stored in the model is used.
sample_cov_keys (list[str] | None (default: None)) – Keys for covariates (batch, etc.) that should also be taken into account when computing the differential abundance. At the moment, only discrete covariates are supported.
sample_subset (list[str] | None (default: None)) – Only computes differential abundance for these sample labels.
compute_log_enrichment (bool (default: False)) – Whether to compute the log enrichment scores for each covariate value.
omit_original_sample (bool (default: True)) – If true, each cell’s sample-of-origin is discarded to compute aggregate posteriors. Only relevant if sample_cov_keys is not None.
batch_size (int (default: 128)) – Minibatch size for computing the differential abundance.

Return type:

Dataset

Returns:

A dataset with data variables:

"log_probs": Array of shape (n_cells, n_samples) containing the log probabilities
for each cell across samples.
"{cov_key}_log_probs": For each key in sample_cov_keys, an array of shape
(n_cells, _cov_values) containing the log probabilities for each cell across covariate values.

TorchMRVI.differential_expression(adata=None, sample_cov_keys=None, sample_subset=None, batch_size=128, use_vmap='auto', normalize_design_matrix=True, add_batch_specific_offsets=False, mc_samples=50, store_lfc=False, store_lfc_metadata_subset=None, store_baseline=False, eps_lfc=0.0001, filter_inadmissible_samples=False, lambd=0.0, delta=0.3, **filter_samples_kwargs)[source]#

Compute cell-specific multivariate differential expression.

For every cell, we first compute all counterfactual cell-state shifts, defined as e_d = z_d - u, where z_d is the latent representation of the cell for sample d and u is the sample-unaware latent representation. Then, we fit a linear model in each cell of the form: e_d = X_d * beta_d + iid gaussian noise.

Parameters:

sample_cov_keys (list[str] | None (default: None)) – List of sample covariates to consider for the multivariate analysis. These keys should be present in adata.obs.
adata (AnnData | None (default: None)) – AnnData object to use for computing the local sample representation. If None, the analysis is performed on all cells in the dataset.
sample_subset (list[str] | None (default: None)) – Optional list of samples to consider for the multivariate analysis. If None, all samples are considered.
batch_size (int (default: 128)) – Batch size to use for computing the local sample representation.
use_vmap (Literal['auto', True, False] (default: 'auto')) – Whether to use vmap for computing the local sample representation.
normalize_design_matrix (bool (default: True)) – Whether to normalize the design matrix.
add_batch_specific_offsets (bool (default: False)) – Whether to offset the design matrix by adding batch-specific offsets to the design matrix. Setting this option to True is recommended when considering multi-site datasets.
mc_samples (int (default: 50)) – How many MC samples should be taken for computing betas.
store_lfc (bool (default: False)) – Whether to store the log-fold changes in the module. Storing log-fold changes is memory-intensive and may require specifying a smaller set of cells to analyze, e.g., by specifying adata.
store_lfc_metadata_subset (list[str] | None (default: None)) – Specifies a subset of metadata for which log-fold changes are computed. These keys must be a subset of sample_cov_keys. Only applies when store_lfc=True.
store_baseline (bool (default: False)) – Whether to store the expression in the module if logfoldchanges are computed.
eps_lfc (float (default: 0.0001)) – Epsilon to add to the log-fold changes to avoid detecting genes with low expression.
filter_inadmissible_samples (bool (default: False)) – Whether to filter out-of-distribution samples prior to performing the analysis.
lambd (float (default: 0.0)) – Regularization parameter for the linear model.
delta (float | None (default: 0.3)) – LFC threshold used to compute posterior DE probabilities. If None does not compute them to save memory consumption.
filter_samples_kwargs – Keyword arguments to pass to get_outlier_cell_sample_pairs().

Return type:

Dataset

Returns:

A dataset containing the results of the differential expression analysis:

"beta": Coefficients for each covariate across cells and latent dimensions.
"effect_size": Effect sizes for each covariate across cells.
"pvalue": P-values for each covariate across cells.
"padj": Adjusted P-values for each covariate across cells using the
Benjamini-Hochberg procedure.
"lfc": Log fold changes for each covariate across cells and genes, if store_lfc
is True.
"lfc_std": Standard deviation of log fold changes, if store_lfc is True and
delta is not None.
"pde": Posterior DE probabilities, if store_lfc is True and delta is not
None.
"baseline_expression": Baseline expression levels for each covariate across cells and
genes, if store_baseline is True.
"n_samples": Number of admissible samples for each cell, if
filter_inadmissible_samples is True.

TorchMRVI.get_aggregated_posterior(adata=None, sample=None, indices=None, batch_size=256)[source]#

Computes the aggregated posterior over the u latent representations.

For the specified samples, it computes the aggregated posterior over the u latent representations. Returns a NumPyro MixtureSameFamily distribution.

Parameters:

adata (AnnData | None (default: None)) – AnnData object to use. Defaults to the AnnData object used to initialize the model.
sample (str | int | None (default: None)) – Name or index of the sample to filter on. If None, uses all cells.
indices (Union[_Buffer, _SupportsArray[dtype[Any]], _NestedSequence[_SupportsArray[dtype[Any]]], complex, bytes, str, _NestedSequence[complex | bytes | str], None] (default: None)) – Indices of cells to use.
batch_size (int (default: 256)) – Batch size to use for computing the latent representation.

Return type:

Distribution

Returns:

A mixture distribution of the aggregated posterior.

TorchMRVI.get_anndata_manager(adata, required=False)[source]#

Retrieves the AnnDataManager for a given AnnData object.

Requires self.id has been set. Checks for an AnnDataManager specific to this model instance.

Parameters:

adata (AnnData | MuData) – AnnData object to find a manager instance for.
required (bool (default: False)) – If True, errors on missing manager. Otherwise, returns None when manager is missing.

Return type:

AnnDataManager | None

TorchMRVI.get_batch_representation(adata=None, indices=None, batch_size=None)[source]#

Get the batch representation for a given set of indices.

Return type:: ndarray

TorchMRVI.get_elbo(adata=None, indices=None, batch_size=None, dataloader=None, return_mean=True, data_loader_kwargs=None, **kwargs)[source]#

Compute the evidence lower bound (ELBO) on the data.

The ELBO is the reconstruction error plus the Kullback-Leibler (KL) divergences between the variational distributions and the priors. It is different from the marginal log-likelihood; specifically, it is a lower bound on the marginal log-likelihood plus a term that is constant with respect to the variational distribution. It still gives good insights on the modeling of the data and is fast to compute.

Parameters:

adata (AnnData | None (default: None)) – AnnData object with var_names in the same order as the ones used to train the model. If None and dataloader is also None, it defaults to the object used to initialize the model.
indices (Sequence[int] | None (default: None)) – Indices of observations in adata to use. If None, defaults to all observations. Ignored if dataloader is not None.
batch_size (int | None (default: None)) – Minibatch size for the forward pass. If None, defaults to scvi.settings.batch_size. Ignored if dataloader is not None.
dataloader (Iterator[dict[str, Tensor | None]] | None (default: None)) – An iterator over minibatches of data on which to compute the metric. The minibatches should be formatted as a dictionary of Tensor with keys as expected by the model. If None, a dataloader is created from adata.
return_mean (bool (default: True)) – Whether to return the mean of the ELBO or the ELBO for each observation.
data_loader_kwargs (dict | None (default: None)) – Keyword args for data loader, in dict form.
**kwargs – Additional keyword arguments to pass into the forward method of the module.

Return type:

float

Returns:

Evidence lower bound (ELBO) of the data.

Notes

This is not the negative ELBO, so higher is better.

TorchMRVI.get_feature_correlation_matrix(adata=None, indices=None, n_samples=10, batch_size=64, rna_size_factor=1000, transform_batch=None, correlation_type='spearman', silent=True)[source]#

Generate gene-gene correlation matrix using scvi uncertainty and expression.

Parameters:

adata (AnnData | None (default: None)) – AnnData object with equivalent structure to initial AnnData. If None, defaults to the AnnData object used to initialize the model.
indices (list[int] | None (default: None)) – Indices of cells in adata to use. If None, all cells are used.
n_samples (int (default: 10)) – Number of posterior samples to use for estimation.
batch_size (int (default: 64)) – Minibatch size for data loading into model. Defaults to scvi.settings.batch_size.
rna_size_factor (int (default: 1000)) – size factor for RNA prior to sampling gamma distribution.
transform_batch (list[int | float | str] | None (default: None)) –
Batches to condition on. If transform_batch is:
- None, then real observed batch is used.
- int, then batch transform_batch is used.
- list of int, then values are averaged over provided batches.
correlation_type (Literal['spearman', 'pearson'] (default: 'spearman')) – One of “pearson”, “spearman”.
%(de_silent)s

Return type:

DataFrame

Returns:

Gene-gene correlation matrix

TorchMRVI.get_from_registry(adata, registry_key)[source]#

Returns the object in AnnData associated with the key in the data registry.

AnnData object should be registered with the model prior to calling this function via the self._validate_anndata method.

Parameters:

registry_key (str) – key of object to get from the data registry.
adata (AnnData | MuData) – AnnData to pull data from.

Return type:

ndarray

Returns:

The requested data as a NumPy array.

TorchMRVI.get_importance_weights(adata, indices, qz, px, zs, max_cells=1024, truncation=False, n_mc_samples=500, n_mc_samples_per_pass=250, **data_loader_kwargs)[source]#

Computes importance weights for the given samples.

This method computes importance weights for every latent code in zs as a way to encourage latent codes providing high likelihoods across many cells in the considered subpopulation.

Parameters:

adata (AnnData | None) – Data to use for computing importance weights.
indices (list[int] | None) – Indices of cells in adata to use.
distributions – Dictionary of distributions associated with indices.
qz (Distribution) – Variational posterior distributions of the cells, aligned with indices.
px (Distribution) – Count distributions of the cells, aligned with indices.
zs (Tensor) – Samples associated with indices.
max_cells (int (default: 1024)) – Maximum number of cells used to estimated the importance weights
truncation (bool (default: False)) – Whether importance weights should be truncated. If True, the importance weights are truncated as described in [Ionides, 2008]. In particular, the provided value is used to threshold importance weights as a way to reduce the variance of the estimator.
n_mc_samples (int (default: 500)) – Number of Monte Carlo samples to use for estimating the importance weights, by default 500
n_mc_samples_per_pass (int (default: 250)) – Number of Monte Carlo samples to use for each pass, by default 250
**data_loader_kwargs – Keyword args for data loader.

Return type:

ndarray

Returns:

importance_weights Numpy array containing importance weights aligned with the provided indices.

Notes

This method assumes a normal prior on the latent space.

TorchMRVI.get_latent_library_size(adata=None, indices=None, give_mean=True, batch_size=None, dataloader=None, **data_loader_kwargs)[source]#

Returns the latent library size for each cell.

This is denoted as \(\ell_n\) in the scVI paper.

Parameters:

adata (AnnData | None (default: None)) – AnnData object with equivalent structure to initial AnnData. If None, defaults to the AnnData object used to initialize the model.
indices (list[int] | None (default: None)) – Indices of cells in adata to use. If None, all cells are used.
give_mean (bool (default: True)) – Return the mean or a sample from the posterior distribution.
batch_size (int | None (default: None)) – Minibatch size for data loading into model. Defaults to scvi.settings.batch_size.
dataloader (Iterator[dict[str, Tensor | None]] | None (default: None)) – An iterator over minibatches of data on which to compute the metric. The minibatches should be formatted as a dictionary of Tensor with keys as expected by the model. If None, a dataloader is created from adata.
**data_loader_kwargs – Keyword args for data loader.

Return type:

ndarray

TorchMRVI.get_latent_representation(adata=None, indices=None, batch_size=None, use_mean=True, give_z=False, dataloader=None)[source]#

Compute the latent representation of the data.

Parameters:

adata (AnnData | None (default: None)) – AnnData object to use. Defaults to the AnnData object used to initialize the model.
indices (Union[_Buffer, _SupportsArray[dtype[Any]], _NestedSequence[_SupportsArray[dtype[Any]]], complex, bytes, str, _NestedSequence[complex | bytes | str], None] (default: None)) – Indices of cells to use.
batch_size (int | None (default: None)) – Batch size to use for computing the latent representation.
use_mean (bool (default: True)) – Whether to use the mean of the distribution as the latent representation.
give_z (bool (default: False)) – Whether to return the z latent representation or the u latent representation.
dataloader (Iterator[dict[str, Tensor | None]] | None (default: None)) – An iterator over minibatches of data on which to compute the metric. The minibatches should be formatted as a dictionary of Tensor with keys as expected by the model. If None, a dataloader is created from adata.

Return type:

ndarray[tuple[Any, ...], dtype[TypeVar(_ScalarT, bound= generic)]]

Returns:

The latent representation of the data.

TorchMRVI.get_likelihood_parameters(adata=None, indices=None, n_samples=1, give_mean=False, batch_size=None, dataloader=None, **data_loader_kwargs)[source]#

Estimates for the parameters of the likelihood \(p(x \mid z)\).

Parameters:

adata (AnnData | None (default: None)) – AnnData object with equivalent structure to initial AnnData. If None, defaults to the AnnData object used to initialize the model.
indices (list[int] | None (default: None)) – Indices of cells in adata to use. If None, all cells are used.
n_samples (int | None (default: 1)) – Number of posterior samples to use for estimation.
give_mean (bool | None (default: False)) – Return expected value of parameters or a samples
batch_size (int | None (default: None)) – Minibatch size for data loading into model. Defaults to scvi.settings.batch_size.
dataloader (Iterator[dict[str, Tensor | None]] | None (default: None)) – An iterator over minibatches of data on which to compute the metric. The minibatches should be formatted as a dictionary of Tensor with keys as expected by the model. If None, a dataloader is created from adata.
**data_loader_kwargs – Keyword args for data loader.

Return type:

dict[str, ndarray]

TorchMRVI.get_local_sample_distances(adata=None, batch_size=256, use_mean=True, normalize_distances=False, use_vmap='auto', groupby=None, keep_cell=True, norm='l2', mc_samples=10)[source]#

Compute local sample distances.

Computes cell-specific distances between samples, of size (n_sample, n_sample), stored as a Dataset, with variable name "cell", of size (n_cell, n_sample, n_sample). If in addition, groupby is provided, distances are also aggregated by group. In this case, the group-specific distances via the group name key.

Parameters:

adata (AnnData | None (default: None)) – AnnData object to use for computing the local sample representation.
batch_size (int (default: 256)) – Batch size to use for computing the local sample representation.
use_mean (bool (default: True)) – Whether to use the mean of the latent representation as the local sample representation.
normalize_distances (bool (default: False)) – Whether to normalize the local sample distances. Normalizes by the standard deviation of the original intra-sample distances. Only works with use_mean=False.
use_vmap (Literal['auto', True, False] (default: 'auto')) – Whether to use vmap for computing the local sample representation. Disabling vmap can be useful if running out of memory on a GPU.
groupby (list[str] | str | None (default: None)) – List of categorical keys or single key of the anndata that is used to group the cells.
keep_cell (bool (default: True)) – Whether to keep the original cell sample-sample distance matrices.
norm (str (default: 'l2')) – Norm to use for computing the local sample distances.
mc_samples (int (default: 10)) – Number of Monte Carlo samples to use for computing the local sample distances. Only relevant if use_mean=False.

Return type:

Dataset

TorchMRVI.get_local_sample_representation(adata=None, indices=None, batch_size=256, use_mean=True, use_vmap='auto')[source]#

Compute the local sample representation of the cells in the adata object.

For each cell, it returns a matrix of size (n_sample, n_features).

Parameters:

adata (AnnData | None (default: None)) – AnnData object to use for computing the local sample representation.
batch_size (int (default: 256)) – Batch size to use for computing the local sample representation.
use_mean (bool (default: True)) – Whether to use the mean of the latent representation as the local sample representation.
use_vmap (Literal['auto', True, False] (default: 'auto')) – Whether to use vmap for computing the local sample representation. Disabling vmap can be useful if running out of memory on a GPU.

Return type:

DataArray

TorchMRVI.get_marginal_ll(adata=None, indices=None, n_mc_samples=1000, batch_size=None, return_mean=True, dataloader=None, data_loader_kwargs=None, **kwargs)[source]#

Compute the marginal log-likehood of the data.

The computation here is a biased estimator of the marginal log-likelihood of the data.

Parameters:

adata (AnnData | None (default: None)) – AnnData object with var_names in the same order as the ones used to train the model. If None and dataloader is also None, it defaults to the object used to initialize the model.
indices (Sequence[int] | None (default: None)) – Indices of observations in adata to use. If None, defaults to all observations. Ignored if dataloader is not None.
n_mc_samples (int (default: 1000)) – Number of Monte Carlo samples to use for the estimator. Passed into the module’s marginal_ll method.
batch_size (int | None (default: None)) – Minibatch size for the forward pass. If None, defaults to scvi.settings.batch_size. Ignored if dataloader is not None.
return_mean (bool (default: True)) – Whether to return the mean of the marginal log-likelihood or the marginal-log likelihood for each observation.
dataloader (Iterator[dict[str, Tensor | None]] (default: None)) – An iterator over minibatches of data on which to compute the metric. The minibatches should be formatted as a dictionary of Tensor with keys as expected by the model. If None, a dataloader is created from adata.
data_loader_kwargs (dict | None (default: None)) – Keyword args for data loader, in dict form.
**kwargs – Additional keyword arguments to pass into the module’s marginal_ll method.

Return type:

float | Tensor

Returns:

If True, returns the mean marginal log-likelihood. Otherwise returns a tensor of shape (n_obs,) with the marginal log-likelihood for each observation.

Notes

This is not the negative log-likelihood, so higher is better.

TorchMRVI.get_normalized_expression(adata=None, indices=None, transform_batch=None, gene_list=None, library_size='latent', n_samples=1, n_samples_overall=None, weights=None, batch_size=None, return_mean=False, return_numpy=None, silent=True, dataloader=None, data_loader_kwargs=None, **importance_weighting_kwargs)[source]#

Returns the normalized (decoded) gene expression.

This is denoted as \(\rho_n\) in the scVI paper.

Parameters:

adata (AnnData | None (default: None)) – AnnData object with equivalent structure to initial AnnData. If None, defaults to the AnnData object used to initialize the model.
indices (list[int] | None (default: None)) – Indices of cells in adata to use. If None, all cells are used.
transform_batch (list[int | float | str] | None (default: None)) – Batch to condition on. If transform_batch is: - None, then the real observed batch is used. - int, then batch transform_batch is used. - Otherwise based on string
gene_list (list[str] | None (default: None)) – Return frequencies of expression for a subset of genes. This can save memory when working with large datasets, and few genes are of interest.
library_size (Union[float, Literal['latent']] (default: 'latent')) – Scale the expression frequencies to a common library size. This allows gene expression levels to be interpreted on a common scale of relevant magnitude. If set to “latent”, use the latent library size.
n_samples (int (default: 1)) – Number of posterior samples to use for estimation.
n_samples_overall (int (default: None)) – Number of posterior samples to use for estimation. Overrides n_samples.
weights (Optional[Literal['uniform', 'importance']] (default: None)) – Weights to use for sampling. If None, defaults to “uniform”.
batch_size (int | None (default: None)) – Minibatch size for data loading into the model. Defaults to scvi.settings.batch_size.
return_mean (bool (default: False)) – Whether to return the mean of the samples.
return_numpy (bool | None (default: None)) – Return a ndarray instead of a DataFrame. DataFrame includes gene names as columns. If either n_samples=1 or return_mean=True, defaults to False. Otherwise, it defaults to True.
%(de_silent)s
dataloader (Iterator[dict[str, Tensor | None]] | None (default: None)) – An iterator over minibatches of data on which to compute the metric. The minibatches should be formatted as a dictionary of Tensor with keys as expected by the model. If None, a dataloader is created from adata.
data_loader_kwargs (dict | None (default: None)) – Keyword args for data loader, in dict form.
importance_weighting_kwargs – Keyword arguments passed into get_importance_weights().

Return type:

ndarray | DataFrame

Returns:

If n_samples is provided and return_mean is False, this method returns a 3d tensor of shape (n_samples, n_cells, n_genes). If n_samples is provided and return_mean is True, it returns a 2d tensor of shape (n_cells, n_genes). In this case, the return type is DataFrame unless return_numpy is True. Otherwise, the method expects n_samples_overall to be provided and returns a 2d tensor of shape (n_samples_overall, n_genes).

TorchMRVI.get_outlier_cell_sample_pairs(adata=None, subsample_size=5000, quantile_threshold=0.05, admissibility_threshold=0.0, batch_size=256)[source]#

Compute admissibility scores for cell-sample pairs.

This function computes the posterior distribution for u for each cell. Then, for every cell, it computes the log-probability of the cell under the posterior of each cell each sample and takes the maximum value for a given sample as a measure of admissibility for that sample. Additionally, it computes a threshold that determines if a cell-sample pair is admissible based on the within-sample admissibility scores.

Parameters:

adata (AnnData | None (default: None)) – AnnData object containing the cells for which to compute the outlier cell-sample pairs.
subsample_size (int (default: 5000)) – Number of cells to use from each sample to approximate the posterior. If None, uses all of the available cells.
quantile_threshold (float (default: 0.05)) – Quantile of the within-sample log probabilities to use as a baseline for admissibility.
admissibility_threshold (float (default: 0.0)) – Threshold for admissibility. Cell-sample pairs with admissibility below this threshold are considered outliers.
batch_size (int (default: 256)) – Size of the batch to use for computing outlier cell-sample pairs.

Return type:

Dataset

TorchMRVI.get_reconstruction_error(adata=None, indices=None, batch_size=None, dataloader=None, return_mean=True, data_loader_kwargs=None, **kwargs)[source]#

Compute the reconstruction error on the data.

The reconstruction error is the negative log likelihood of the data given the latent variables. It is different from the marginal log-likelihood, but still gives good insights on the modeling of the data and is fast to compute. This is typically written as \(p(x \mid z)\), the likelihood term given one posterior sample.

Parameters:

adata (AnnData | None (default: None)) – AnnData object with var_names in the same order as the ones used to train the model. If None and dataloader is also None, it defaults to the object used to initialize the model.
indices (Sequence[int] | None (default: None)) – Indices of observations in adata to use. If None, defaults to all observations. Ignored if dataloader is not None
batch_size (int | None (default: None)) – Minibatch size for the forward pass. If None, defaults to scvi.settings.batch_size. Ignored if dataloader is not None
dataloader (Iterator[dict[str, Tensor | None]] | None (default: None)) – An iterator over minibatches of data on which to compute the metric. The minibatches should be formatted as a dictionary of Tensor with keys as expected by the model. If None, a dataloader is created from adata.
return_mean (bool (default: True)) – Whether to return the mean reconstruction loss or the reconstruction loss for each observation.
data_loader_kwargs (dict | None (default: None)) – Keyword args for data loader, in dict form.
**kwargs – Additional keyword arguments to pass into the forward method of the module.

Return type:

dict[str, float]

Returns:

Reconstruction error for the data.

Notes

This is not the negative reconstruction error, so higher is better.

TorchMRVI.get_setup_arg(setup_arg)[source]#

Returns the string provided to setup of a specific setup_arg.

Return type:: attrdict

TorchMRVI.get_state_registry(registry_key)[source]#

Returns the state registry for the AnnDataField registered with this instance.

Return type:: attrdict

TorchMRVI.get_var_names(legacy_mudata_format=False)[source]#

Variable names of input data.

Return type:: dict

classmethod TorchMRVI.load(dir_path, adata=None, accelerator='auto', device='auto', prefix=None, backup_url=None, datamodule=None, allowed_classes_names_list=None)[source]#

Instantiate a model from the saved output.

Parameters:

dir_path (str) – Path to saved outputs.
adata (AnnData | MuData | None (default: None)) – AnnData organized in the same way as data used to train model. It is not necessary to run setup_anndata, as AnnData is validated against the saved scvi setup dictionary. If None, will check for and load anndata saved with the model. If False, will load the model without AnnData.
accelerator (str (default: 'auto')) – Supports passing different accelerator types (“cpu”, “gpu”, “tpu”, “ipu”, “hpu”, “mps, “auto”) as well as custom accelerator instances.
device (int | str (default: 'auto')) – The device to use. Can be set to a non-negative index (int or str) or “auto” for automatic selection based on the chosen accelerator. If set to “auto” and accelerator is not determined to be “cpu”, then device will be set to the first available device.
prefix (str | None (default: None)) – Prefix of saved file names.
backup_url (str | None (default: None)) – URL to retrieve saved outputs from if not present on disk.
datamodule (LightningDataModule | None (default: None)) – EXPERIMENTAL A LightningDataModule instance to use for training in place of the default DataSplitter. Can only be passed in if the model was not initialized with AnnData.
allowed_classes_names_list (list[str] | None (default: None)) – list of allowed classes names to be loaded (besides the original class name)

Returns:

Model with loaded state dictionaries.

Examples

>>> model = ModelClass.load(save_path, adata)
>>> model.get_....

classmethod TorchMRVI.load_query_data(adata=None, reference_model=None, registry=None, inplace_subset_query_vars=False, accelerator='auto', device='auto', unfrozen=False, freeze_dropout=False, freeze_expression=True, freeze_decoder_first_layer=True, freeze_batchnorm_encoder=True, freeze_batchnorm_decoder=False, freeze_classifier=True, transfer_batch=True, datamodule=None)[source]#

Online update of a reference model with scArches algorithm [Lotfollahi et al., 2021].

Parameters:

adata (AnnData | MuData (default: None)) – AnnData organized in the same way as data used to train model. It is not necessary to run setup_anndata, as AnnData is validated against the registry.
reference_model (str | BaseModelClass (default: None)) – Either an already instantiated model of the same class or a path to saved outputs for the reference model.
inplace_subset_query_vars (bool (default: False)) – Whether to subset and rearrange query vars inplace based on vars used to train the reference model.
accelerator (str (default: 'auto')) – Supports passing different accelerator types (“cpu”, “gpu”, “tpu”, “ipu”, “hpu”, “mps, “auto”) as well as custom accelerator instances.
device (int | str (default: 'auto')) – The device to use. Can be set to a non-negative index (int or str) or “auto” for automatic selection based on the chosen accelerator. If set to “auto” and accelerator is not determined to be “cpu”, then device will be set to the first available device.
unfrozen (bool (default: False)) – Override all other freeze options for a fully unfrozen model
freeze_dropout (bool (default: False)) – Whether to freeze dropout during training
freeze_expression (bool (default: True)) – Freeze neurons corresponding to expression in first layer
freeze_decoder_first_layer (bool (default: True)) – Freeze neurons corresponding to first layer in decoder
freeze_batchnorm_encoder (bool (default: True)) – Whether to freeze batchnorm weight and bias during training for encoder
freeze_batchnorm_decoder (bool (default: False)) – Whether to freeze batchnorm weight and bias during training for decoder
freeze_classifier (bool (default: True)) – Whether to freeze classifier completely. Only applies to SCANVI.
transfer_batch (bool (default: True)) – Allow for surgery on the batch covariate. Only applies to SYSVI.
datamodule (LightningDataModule | None (default: None)) – EXPERIMENTAL A LightningDataModule instance to use for training in place of the default DataSplitter. Can only be passed in if the model was not initialized with AnnData.

static TorchMRVI.load_registry(dir_path, prefix=None)[source]#

Return the full registry saved with the model.

Parameters:

dir_path (str) – Path to saved outputs.
prefix (str | None (default: None)) – Prefix of saved file names.

Return type:

dict

Returns:

The full registry saved with the model

TorchMRVI.minify_adata(minified_data_type='latent_posterior_parameters', use_latent_qzm_key='X_latent_qzm', use_latent_qzv_key='X_latent_qzv')[source]#

Minify the model’s adata.

Minifies the AnnData object associated with the model according to the method specified by minified_data_type and registers the new fields with the model’s AnnDataManager. This also sets the minified_data_type attribute of the underlying BaseModuleClass instance.

Parameters:

minified_data_type (Literal['latent_posterior_parameters'] (default: 'latent_posterior_parameters')) –
Method for minifying the data. One of the following:
- "latent_posterior_parameters": Store the latent posterior mean and variance in
  obsm using the keys use_latent_qzm_key and use_latent_qzv_key.
- "latent_posterior_parameters_with_counts": Store the latent posterior mean and
  variance in obsm using the keys use_latent_qzm_key and use_latent_qzv_key, and the raw count data in X.
use_latent_qzm_key (str (default: 'X_latent_qzm')) – Key to use for storing the latent posterior mean in obsm when minified_data_type is "latent_posterior".
use_latent_qzv_key (str (default: 'X_latent_qzv')) – Key to use for storing the latent posterior variance in obsm when minified_data_type is "latent_posterior".

Return type:

None

Notes

The modification is not done inplace – instead the model is assigned a new (minified) version of the AnnData.

TorchMRVI.posterior_predictive_sample(adata=None, indices=None, transform_batch=None, n_samples=1, gene_list=None, batch_size=None, dataloader=None, silent=True, **data_loader_kwargs)[source]#

Generate predictive samples from the posterior predictive distribution.

The posterior predictive distribution is denoted as \(p(\hat{x} \mid x)\), where \(x\) is the input data and \(\hat{x}\) is the sampled data.

We sample from this distribution by first sampling n_samples times from the posterior distribution \(q(z \mid x)\) for a given observation, and then sampling from the likelihood \(p(\hat{x} \mid z)\) for each of these.

Parameters:

adata (AnnData | None (default: None)) – AnnData object with an equivalent structure to the model’s dataset. If None, defaults to the AnnData object used to initialize the model.
indices (list[int] | None (default: None)) – Indices of the observations in adata to use. If None, defaults to all the observations.
transform_batch (list[int | float | str] | None (default: None)) – Batch to condition on. If transform_batch is: - None, then real observed batch is used. - int, then batch transform_batch is used. - Otherwise based on string
n_samples (int (default: 1)) – Number of Monte Carlo samples to draw from the posterior predictive distribution for each observation.
gene_list (list[str] | None (default: None)) – Names of the genes to which to subset. If None, defaults to all genes.
batch_size (int | None (default: None)) – Minibatch size to use for data loading and model inference. Defaults to scvi.settings.batch_size. Passed into BaseModelClass._make_data_loader.
dataloader (Iterator[dict[str, Tensor | None]] | None (default: None)) – An iterator over minibatches of data on which to compute the metric. The minibatches should be formatted as a dictionary of Tensor with keys as expected by the model. If None, a dataloader is created from adata.
**data_loader_kwargs – Keyword args for data loader.

Return type:

GCXS

Returns:

Sparse multidimensional array of shape (n_obs, n_vars) if n_samples == 1, else (n_obs, n_vars, n_samples).

static TorchMRVI.prepare_query_anndata(adata, reference_model, return_reference_var_names=False, inplace=True)[source]#

Prepare data for query integration.

This function will return a new AnnData object with padded zeros for missing features, as well as correctly sorted features.

Parameters:

adata (AnnData) – AnnData organized in the same way as data used to train model. It is not necessary to run setup_anndata, as AnnData is validated against the registry.
reference_model (str | BaseModelClass) – Either an already instantiated model of the same class or a path to saved outputs for the reference model.
return_reference_var_names (bool (default: False)) – Only load and return reference var names if True.
inplace (bool (default: True)) – Whether to subset and rearrange query vars inplace or return new AnnData.

Return type:

AnnData | Index | None

Returns:

Query adata ready to use in load_query_data unless return_reference_var_names in which case a pd.Index of reference var names is returned.

static TorchMRVI.prepare_query_mudata(mdata, reference_model, return_reference_var_names=False, inplace=True)[source]#

Prepare multimodal dataset for query integration.

This function will return a new MuData object such that the AnnData objects for individual modalities are given padded zeros for missing features, as well as correctly sorted features.

Parameters:

mdata (MuData) – MuData organized in the same way as data used to train the model. It is not necessary to run setup_mudata, as MuData is validated against the registry.
reference_model (str | BaseModelClass) – Either an already instantiated model of the same class or a path to saved outputs for the reference model.
return_reference_var_names (bool (default: False)) – Only load and return reference var names if True.
inplace (bool (default: True)) – Whether to subset and rearrange query vars inplace or return new MuData.

Return type:

MuData | dict[str, Index] | None

Returns:

Query mudata ready to use in load_query_data unless return_reference_var_names in which case a dictionary of pd.Index of reference var names is returned.

classmethod TorchMRVI.register_manager(adata_manager)[source]#

Registers an AnnDataManager instance with this model class.

Stores the AnnDataManager reference in a class-specific manager store. Intended for use in the setup_anndata() class method followed up by retrieval of the AnnDataManager via the _get_most_recent_anndata_manager() method in the model init method.

Notes

Subsequent calls to this method with an AnnDataManager instance referring to the same underlying AnnData object will overwrite the reference to previous AnnDataManager.

TorchMRVI.save(dir_path, prefix=None, overwrite=False, save_anndata=False, save_kwargs=None, legacy_mudata_format=False, datamodule=None, **anndata_write_kwargs)[source]#

Save the state of the model.

Neither the trainer optimizer state nor the trainer history are saved. Model files are not expected to be reproducibly saved and loaded across versions until we reach version 1.0.

Parameters:

dir_path (str) – Path to a directory.
prefix (str | None (default: None)) – Prefix to prepend to saved file names.
overwrite (bool (default: False)) – Overwrite existing data or not. If False and directory already exists at dir_path, an error will be raised.
save_anndata (bool (default: False)) – If True, also saves the anndata
save_kwargs (dict | None (default: None)) – Keyword arguments passed into save().
legacy_mudata_format (bool (default: False)) – If True, saves the model var_names in the legacy format if the model was trained with a MuData object. The legacy format is a flat array with variable names across all modalities concatenated, while the new format is a dictionary with keys corresponding to the modality names and values corresponding to the variable names for each modality.
datamodule (LightningDataModule | None (default: None)) – EXPERIMENTAL A LightningDataModule instance to use for training in place of the default DataSplitter. Can only be passed in if the model was not initialized with AnnData.
anndata_write_kwargs – Kwargs for write()

classmethod TorchMRVI.setup_anndata(adata, layer=None, sample_key=None, batch_key=None, labels_key=None, **kwargs)[source]#

Sets up the AnnData object for this model.

A mapping will be created between data fields used by this model to their respective locations in adata. None of the data in adata are modified. Only adds fields to adata.

Parameters:

adata (AnnData) – AnnData object. Rows represent cells, columns represent features.
layer (str | None (default: None)) – if not None, uses this as the key in adata.layers for raw count data.
sample_key (str | None (default: None)) – key in adata.obs for sample information. Categories will automatically be converted into integer categories and saved to adata.obs[‘_scvi_sample’]. If None, assigns the same sample to all the data.
batch_key (str | None (default: None)) – key in adata.obs for batch information. Categories will automatically be converted into integer categories and saved to adata.obs[‘_scvi_batch’]. If None, assigns the same batch to all the data.
labels_key (str | None (default: None)) – key in adata.obs for label information. Categories will automatically be converted into integer categories and saved to adata.obs[‘_scvi_labels’]. If None, assigns the same label to all the data.
**kwargs – Additional keyword arguments passed into register_fields().

TorchMRVI.to_device(device)[source]#

Move the model to the device.

Parameters:: device (str | int | device) – Device to move model to. Options: ‘cpu’ for CPU, integer GPU index (e.g., 0), ‘cuda:X’ where X is the GPU index (e.g. ‘cuda:0’), or a torch.device object (including XLA devices for TPU). See torch.device for more info.

Examples

>>> adata = scvi.data.synthetic_iid()
>>> model = scvi.model.SCVI(adata)
>>> model.to_device("cpu")  # moves model to CPU
>>> model.to_device("cuda:0")  # moves model to GPU 0
>>> model.to_device(0)  # also moves model to GPU 0

TorchMRVI.train(max_epochs=None, accelerator='auto', devices='auto', train_size=None, validation_size=None, batch_size=128, early_stopping=False, plan_kwargs=None, **trainer_kwargs)[source]#

Train the model.

Parameters:

max_epochs (int | None (default: None)) – Maximum number of epochs to train the model. The actual number of epochs may be less if early stopping is enabled. If None, defaults to a heuristic based on get_max_epochs_heuristic().
accelerator (str | None (default: 'auto')) – Supports passing different accelerator types (“cpu”, “gpu”, “tpu”, “ipu”, “hpu”, “mps, “auto”) as well as custom accelerator instances.
devices (int | list[int] | str (default: 'auto')) – The devices to use. Can be set to a non-negative index (int or str), a sequence of device indices (list or comma-separated str), the value -1 to indicate all available devices, or “auto” for automatic selection based on the chosen accelerator. If set to “auto” and accelerator is not determined to be “cpu”, then devices will be set to the first available device.
train_size (float | None (default: None)) – Size of the training set in the range [0.0, 1.0].
validation_size (float | None (default: None)) – Size of the validation set. If None, defaults to 1 - train_size. If train_size + validation_size < 1, the remaining cells belong to a test set.
batch_size (int (default: 128)) – Minibatch size to use during training.
early_stopping (bool (default: False)) – Perform early stopping. Additional arguments can be passed in through **kwargs. See Trainer for further options.
plan_kwargs (dict | None (default: None)) – Additional keyword arguments passed into TrainingPlan.
**trainer_kwargs – Additional keyword arguments passed into Trainer.

TorchMRVI.transfer_fields(adata, **kwargs)[source]#

Transfer fields from a model to an AnnData object.

Return type:: AnnData

TorchMRVI.update_sample_info(adata)[source]#

Initialize/update metadata in the case where additional covariates are added.

Parameters:: adata – AnnData object to update the sample info with. Typically, this corresponds to the working dataset, where additional sample-specific covariates have been added.

Examples

>>> import scanpy as sc
>>> from scvi.external import MRVI
>>> MRVI.setup_anndata(adata, sample_key="sample_id", backend="torch")
>>> model = MRVI(adata)
>>> model.train()
>>> # Update sample info with new covariates
>>> sample_mapper = {"sample_1": "healthy", "sample_2": "disease"}
>>> adata.obs["disease_status"] = adata.obs["sample_id"].map(sample_mapper)
>>> model.update_sample_info(adata)

TorchMRVI.update_setup_method_args(setup_method_args)[source]#

Update setup method args.

Parameters:: setup_method_args (dict) – This is a bit of a misnomer, this is a dict representing kwargs of the setup method that will be used to update the existing values in the registry of this instance.

TorchMRVI.view_anndata_setup(adata=None, hide_state_registries=False)[source]#

Print summary of the setup for the initial AnnData or a given AnnData object.

Parameters:

adata (AnnData | MuData | None (default: None)) – AnnData object setup with setup_anndata or transfer_fields().
hide_state_registries (bool (default: False)) – If True, prints a shortened summary without details of each state registry.

Return type:

None

TorchMRVI.view_registry(hide_state_registries=False)[source]#

Prints summary of the registry.

Parameters:: hide_state_registries (bool (default: False)) – If True, prints a shortened summary without details of each state registry.
Return type:: None

static TorchMRVI.view_setup_args(dir_path, prefix=None)[source]#

Print args used to setup a saved model.

Parameters:

dir_path (str) – Path to saved outputs.
prefix (str | None (default: None)) – Prefix of saved file names.

Return type:

None

TorchMRVI.view_setup_method_args()[source]#

Prints setup kwargs used to produce a given registry.

Parameters:: registry – Registry produced by an AnnDataManager.
Return type:: None

scvi.external.TorchMRVI

Contents

scvi.external.TorchMRVI#

Attributes table#

Methods table#

Attributes#

Methods#