Select Cells for Downstream PC Regression Analysis

This function selects a subset of cells from a Seurat object for downstream analysis based on their principal component (PC) embeddings. The selection ensures that the subset represents the distribution of the full dataset while maintaining computational efficiency.

Usage

select.cell(seurat.obj, p = 0.1, n = NULL, all = FALSE, ...)

Arguments

seurat.obj: A Seurat object that has been processed with prepare.graph. The PC embeddings and graph structure must be stored in the object's NNet.setting.
p: A numeric value between 0 and 1, specifying the proportion of cells to select. If n is provided, p is ignored. Default is 0.1.
n: An integer specifying the total number of cells to select. If NULL, the number is calculated as ceiling(p * total cells). Default is NULL.
all: A logical value indicating whether to select all cells. If TRUE, all cells are used, and no subset selection is performed. Default is FALSE.
...: Additional parameters passed to kmeans for clustering.

Value

A Seurat object with updated NNet.setting, where the selected cells are stored as a vector of indexes named by cell barcodes in the cells field of the NNet.setting list in the misc slot.

Details

This function provides flexibility in selecting cells for downstream PC regression analysis:

Subset Selection: If all = FALSE, a subset of cells is selected based on k-means clustering in the PC space. This ensures that the selected cells represent the overall distribution of the data.

Balanced Sampling: The function identifies cluster centers using k-means and selects the nearest neighbors to achieve a balanced representation.

Full Dataset Usage: If all = TRUE, no cells are excluded, and the entire dataset is used.

By default, the function selects 10 users can provide n to select an exact number of cells or use all = TRUE to include all cells.

Examples

# Assuming `seurat.obj` is a Seurat object containing `NNet.setting` in its `misc` slot.

# Select 10% of the cells
seurat.obj <- select.cell(seurat.obj, p = 0.1)
#> Error: object 'seurat.obj' not found

# Select a fixed number of cells
seurat.obj <- select.cell(seurat.obj, n = 500)
#> Error: object 'seurat.obj' not found

# Use all cells for downstream analysis
seurat.obj <- select.cell(seurat.obj, all = TRUE)
#> Error: object 'seurat.obj' not found