workflows/annotation/scgpt_integration_knn

Description

Cell type annotation workflow that performs scGPT integration of reference and query dataset followed by KNN label transfer.

Type

nextflow_script

Links

Repository

Contributors

Launch

Run this component

Run the following command to execute this component with Nextflow:

cat > params.yaml <<'EOM'  
id: "run"  
modality: [ "rna" ]  
pad_token: [ "<pad>" ]  
pad_value: [ -2 ]  
n_hvg: [ 1200 ]  
dsbn: [ true ]  
batch_size: [ 64 ]  
n_input_bins: [ 51 ]  
leiden_resolution: [ 1 ]  
weights: [ "uniform" ]  
n_neighbors: [ 15 ]  
output: "$id.$key.output.h5mu"  
output_obsm_integrated: [ "X_integrated_scgpt" ]  
publish_dir: "output/"  
EOM

nextflow run https://packages.viash-hub.com/vsh/openpipeline.git \  
  -revision 2.0.0 \  
  -main-script target/nextflow/workflows/annotation/scgpt_integration_knn/main.nf \  
  -params-file params.yaml

Query Input

Name	Type & Properties	Description
--id	string required	ID of the sample.
--input	file required	Input dataset consisting of the (unlabeled) query observations. The dataset is expected to be pre-processed in the same way as --reference.
--modality	string	Which modality to process. Should match the modality of the --reference dataset.
--input_layer	string	Mudata layer (key from layers) to use as input data for scGPT integration; if not specified, X is used. Should match the layer name of the reference dataset.
--input_var_gene_names	string	The .var field in the input (query) dataset containing gene names; if not provided, the .var index will be used.
--input_obs_batch_label	string required	The .obs field in the input (query) dataset containing the batch labels.
--overwrite_existing_key	boolean_true	If provided, will overwrite existing fields in the input dataset when data are copied during the reference alignment process.

Name

Type & Properties

--id

string

required

--input

file

required

--modality

string

--input_layer

string

--input_var_gene_names

string

--input_obs_batch_label

string

required

--overwrite_existing_key

boolean_true

Reference input

Name	Type & Properties	Description
--reference	file required	Reference dataset consisting of observations with cell type labels present in the .obs --reference_obs_batch_label column to train the classifier on. The dataset is expected to be pre-processed in the same way as the --input query dataset(s).
--reference_obs_targets	string required multiple	The `.obs` key(s) of the target labels to tranfer.
--reference_var_gene_names	string	The .var field in the reference dataset containing gene names; if not provided, the .var index will be used.
--reference_obs_batch_label	string required	The .obs field in the reference dataset containing the batch labels.

Name

Type & Properties

--reference

file

required

--reference_obs_targets

string

required

multiple

--reference_var_gene_names

string

--reference_obs_batch_label

string

required

scGPT model

Name	Type & Properties	Description
--model	file required	The scGPT model file. Can either be a foundation model or a fine-tuned model. If the model file is a fine-tuned model, it must contain a key for the checkpoints (--finetuned_checkpoints_key).
--model_vocab	file required	The scGPT model vocabulary file.
--model_config	file required	The scGPT model configuration file.
--finetuned_checkpoints_key	string	Key in the model file containing the pretrained checkpoints. Must be provided when `--model` is a fine-tuned model.

Name

Type & Properties

--model

file

required

--model_vocab

file

required

--model_config

file

required

--finetuned_checkpoints_key

string

Padding arguments

Name	Type & Properties	Description
--pad_token	string	Token used for padding.
--pad_value	integer	The value of the padding token.

Name

Type & Properties

--pad_token

string

--pad_value

integer

HVG subset arguments

Name	Type & Properties	Description
--n_hvg	integer	Number of highly variable genes to subset for.

Name

Type & Properties

--n_hvg

integer

Tokenization arguments

Name	Type & Properties	Description
--max_seq_len	integer	The maximum sequence length of the tokenized data. Defaults to the number of features if not provided.

Name

Type & Properties

--max_seq_len

integer

Embedding arguments

Name	Type & Properties	Description
--dsbn	boolean	Apply domain-specific batch normalization
--batch_size	integer	The batch size to be used for embedding inference.

Name

Type & Properties

--dsbn

boolean

--batch_size

integer

Binning arguments

Name	Type & Properties	Description
--n_input_bins	integer	The number of bins to discretize the data into; When no value is provided, data won't be binned.
--seed	integer	Seed for random number generation used for binning. If not set, no seed is used.

Name

Type & Properties

--n_input_bins

integer

--seed

integer

Leiden clustering options

Name	Type & Properties	Description
--leiden_resolution	double multiple	Control the coarseness of the clustering. Higher values lead to more clusters.

Name

Type & Properties

--leiden_resolution

double

multiple

Neighbor classifier arguments

Name	Type & Properties	Description
--weights	string	Weight function used in prediction. Possible values are: `uniform` (all points in each neighborhood are weighted equally) or `distance` (weight points by the inverse of their distance)
--n_neighbors	integer	The number of neighbors to use in k-neighbor graph structure used for fast approximate nearest neighbor search with PyNNDescent. Larger values will result in more accurate search results at the cost of computation time.

Name

Type & Properties

--weights

string

--n_neighbors

integer

Outputs

Name	Type & Properties	Description
--output	file required output	The query data in .h5mu format with predicted labels predicted from the classifier trained on the reference.
--output_obs_predictions	string multiple	In which `.obs` slots to store the predicted information. If provided, must have the same length as `--reference_obs_targets`. If empty, will default to the `reference_obs_targets` combined with the `"_pred"` suffix.
--output_obs_probability	string multiple	In which `.obs` slots to store the probability of the predictions. If provided, must have the same length as `--reference_obs_targets`. If empty, will default to the `reference_obs_targets` combined with the `"_probability"` suffix.
--output_obsm_integrated	string	In which .obsm slot to store the integrated embedding.
--output_compression	string	The compression format to be used on the output h5mu object.

Name

Type & Properties

--output

file

required

output

--output_obs_predictions

string

multiple

--output_obs_probability

string

multiple

--output_obsm_integrated

string

--output_compression

string