integrate/scvi

Description

Performs scvi integration as done in the human lung cell atlas https://github.com/LungCellAtlas/HLCA

Type

python_script

Contributors

Run this component

Run the following command to execute this component with Nextflow:

cat > params.yaml <<'EOM'  
modality: [ "rna" ]  
obs_batch: [ "sample_id" ]  
output: "$id.$key.output"  
output_model: "$id.$key.output_model"  
obsm_output: [ "X_scvi_integrated" ]  
n_hidden_nodes: [ 128 ]  
n_dimensions_latent_space: [ 30 ]  
n_hidden_layers: [ 2 ]  
dropout_rate: [ 0.1 ]  
dispersion: [ "gene" ]  
gene_likelihood: [ "nb" ]  
use_layer_normalization: [ "both" ]  
use_batch_normalization: [ "none" ]  
early_stopping_monitor: [ "elbo_validation" ]  
early_stopping_patience: [ 45 ]  
early_stopping_min_delta: [ 0 ]  
reduce_lr_on_plateau: [ true ]  
lr_factor: [ 0.6 ]  
lr_patience: [ 30 ]  
n_obs_min_count: [ 0 ]  
n_var_min_count: [ 0 ]  
id: "run"  
publish_dir: "output/"  
EOM

nextflow run https://packages.viash-hub.com/vsh/openpipeline.git \  
  -revision 1.0.2 \  
  -main-script target/nextflow/integrate/scvi/main.nf \  
  -params-file params.yaml

Inputs

Name	Type & Properties	Description
--input -i	file required	Input h5mu file
--modality	string
--input_layer	string	Input layer to use. If None, X is used
--obs_batch	string	Column name discriminating between your batches.
--var_input	string	.var column containing highly variable genes. By default, do not subset genes.
--obs_labels	string	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.
--obs_size_factor	string	Key in adata.obs for size factor information. Instead of using library size as a size factor, the provided size factor column will be used as offset in the mean of the likelihood. Assumed to be on linear scale.
--obs_categorical_covariate	string multiple	Keys in adata.obs that correspond to categorical data. These covariates can be added in addition to the batch covariate and are also treated as nuisance factors (i.e., the model tries to minimize their effects on the latent space). Thus, these should not be used for biologically-relevant factors that you do _not_ want to correct for.
--obs_continuous_covariate	string multiple	Keys in adata.obs that correspond to continuous data. These covariates can be added in addition to the batch covariate and are also treated as nuisance factors (i.e., the model tries to minimize their effects on the latent space). Thus, these should not be used for biologically-relevant factors that you do _not_ want to correct for.

Name

Type & Properties

--input

-i

file

required

--modality

string

--input_layer

string

--obs_batch

string

--var_input

string

--obs_labels

string

--obs_size_factor

string

--obs_categorical_covariate

string

multiple

--obs_continuous_covariate

string

multiple

Outputs

Name	Type & Properties	Description
--output -o	file required output	Output h5mu file.
--output_model	file output	Folder where the state of the trained model will be saved to.
--output_compression	string	The compression format to be used on the output h5mu object.
--obsm_output	string	In which .obsm slot to store the resulting integrated embedding.

Name

Type & Properties

--output

-o

file

required

output

--output_model

file

output

--output_compression

string

--obsm_output

string

SCVI options

Name	Type & Properties	Description
--n_hidden_nodes	integer	Number of nodes per hidden layer.
--n_dimensions_latent_space	integer	Dimensionality of the latent space.
--n_hidden_layers	integer	Number of hidden layers used for encoder and decoder neural-networks.
--dropout_rate	double	Dropout rate for the neural networks.
--dispersion	string	Set the behavior for the dispersion for negative binomial distributions: - gene: dispersion parameter of negative binomial is constant per gene across cells - gene-batch: dispersion can differ between different batches - gene-label: dispersion can differ between different labels - gene-cell: dispersion can differ for every gene in every cell
--gene_likelihood	string	Model used to generate the expression data from a count-based likelihood distribution. - nb: Negative binomial distribution - zinb: Zero-inflated negative binomial distribution - poisson: Poisson distribution

Name

Type & Properties

--n_hidden_nodes

integer

--n_dimensions_latent_space

integer

--n_hidden_layers

integer

--dropout_rate

double

--dispersion

string

--gene_likelihood

string

Variational auto-encoder model options

Name	Type & Properties	Description
--use_layer_normalization	string	Neural networks for which to enable layer normalization.
--use_batch_normalization	string	Neural networks for which to enable batch normalization.
--encode_covariates	boolean_false	Whether to concatenate covariates to expression in encoder
--deeply_inject_covariates	boolean_true	Whether to concatenate covariates into output of hidden layers in encoder/decoder. This option only applies when n_layers > 1. The covariates are concatenated to the input of subsequent hidden layers.
--use_observed_lib_size	boolean_true	Use observed library size for RNA as scaling factor in mean of conditional distribution.

Name

Type & Properties

--use_layer_normalization

string

--use_batch_normalization

string

--encode_covariates

boolean_false

--deeply_inject_covariates

boolean_true

--use_observed_lib_size

boolean_true

Early stopping arguments

Name	Type & Properties	Description
--early_stopping	boolean	Whether to perform early stopping with respect to the validation set.
--early_stopping_monitor	string	Metric logged during validation set epoch.
--early_stopping_patience	integer	Number of validation epochs with no improvement after which training will be stopped.
--early_stopping_min_delta	double	Minimum change in the monitored quantity to qualify as an improvement, i.e. an absolute change of less than min_delta, will count as no improvement.

Name

Type & Properties

--early_stopping

boolean

--early_stopping_monitor

string

--early_stopping_patience

integer

--early_stopping_min_delta

double

Learning parameters

Name	Type & Properties	Description
--max_epochs	integer	Number of passes through the dataset, defaults to (20000 / number of cells) * 400 or 400; whichever is smallest.
--reduce_lr_on_plateau	boolean	Whether to monitor validation loss and reduce learning rate when validation set `lr_scheduler_metric` plateaus.
--lr_factor	double	Factor to reduce learning rate.
--lr_patience	double	Number of epochs with no improvement after which learning rate will be reduced.

Name

Type & Properties

--max_epochs

integer

--reduce_lr_on_plateau

boolean

--lr_factor

double

--lr_patience

double

Data validition

Name	Type & Properties	Description
--n_obs_min_count	integer	Minimum number of cells threshold ensuring that every obs_batch category has sufficient observations (cells) for model training.
--n_var_min_count	integer	Minimum number of genes threshold ensuring that every var_input filter has sufficient observations (genes) for model training.

Name

Type & Properties

--n_obs_min_count

integer

--n_var_min_count

integer