Analysis steps for Harmony integration — harmony

This function implements all the analysis steps for performing Harmony integration on a Seurat object. These include, 1. merge all samples in a single Seurat object (if a list of Seurat objects is provided) 2. data normalisation, 3. identification of HVGs, 4. Scaling of expression values and computing PCA, 5. running Harmony integration, 8. Generating plots.

Usage

harmony_analysis(
  seu,
  batch_id = "sample",
  npcs,
  dims.use = NULL,
  plot_dir = NULL,
  n_hvgs = 3000,
  max.iter.harmony = 50,
  seed = 1,
  fig.res = 200,
  ...
)

Arguments

seu: Seurat object (required).
batch_id: Name of batch to try and remove with data integration (required). Can also be a vector if multiple batch information are present. Should be a column name in Seurat 'meta.data'. Default is "sample". This parameter is called 'group.by.vars' in Harmony.
npcs: Number of principal components.
dims.use: Vector with PCs to use for Harmony integration, e.g. 1:50.
plot_dir: Directory to save generated plots. If NULL, plots are not saved.
n_hvgs: Number of highly variable genes (HVGs) to compute, which will be used as input to PCA.
max.iter.harmony: Maximum number of iterations for Harmony integration.
seed: Set a random seed, for reproducibility.
fig.res: Figure resolution in ppi (see 'png' function).
...: Additional named parameters passed to RunHarmony and other Seurat processing functions, such as RunPCA and ScaleData.

Value

Updated Seurat object with integrated and processed data.

Author

C.A.Kapourani C.A.Kapourani@ed.ac.uk