Quantify batch effect

Understanding the existing batch effect in your data is crucial for achieving a good integration.

Using principle

example_batch_analysis:
  input:
    preprocessing:
      custom_file_name: custom/file/path.zarr # file path
    batch_analysis: preprocessing
  batch_analysis:
    sample: sample # smallest entity of a batch, e.g. <bio_sample>-<pool> if there is no 1-1 matching of sample to pool
    n_permutations: 100 # you have small data, so should be quick
    covariates:
      - # obs columns of potential batch covariates
  preprocessing:
    highly_variable_genes:
      n_top_genes: 2000
      # no batch_key bc we want to maximise batch effect
    assemble: # only pca is needed here
      - pca
    colors: # obs columns that you want in your UMAP
      - cell_type