This pipeline performs a grid search to optimize hyperparameters for scVI
. The grid is constructed by combining elements from predefined lists of model input parameters. After running scVI
with each set of parameters, integration metrics are calculated using scib-metrics
.
main.nf
– The Nextflow pipeline that orchestrates the entire workflow.nextflow.config
– Configuration file that:- Submits processes to LSF on Sanger's HPC.
- Ensures the correct environment is set via a Singularity container (absolute path).
- Defines global default parameters.
examples/inputs.py
– Example.py
file specifying the path toanndata
and configurations forscVI
andscib-metrics
.examples/run_scVI_metrics.sh
– Example script to execute the pipeline.Dockerfile
– Defines a Docker image withscVI
andscib-metrics
version0.2.2
.
--input_file
– Path to a.py
file specifying:- The path to
anndata
. - Configuration details for
scVI
andscib-metrics
.
- The path to
--help
— Display this help message--umap
— Set this flag to calculate umap