colDeltaCor_all

This function calculates cell-to-cell correlations based on observed and perturbed gene expression matrices, with an option to limit computations to pairs of cells connected in a KNN graph. Correlations are computed in parallel to optimize performance.

Usage

colDeltaCor_knn(
  observed_matrix,
  delta_matrix,
  knn_graph = NULL,
  n_cores = detectCores() - 1
)

Arguments

observed_matrix

A matrix of observed gene expression values (genes by cells).

delta_matrix

A matrix of perturbed gene expression values (genes by cells).

knn_graph

Optional; an adjacency matrix representing the k-nearest-neighbor (KNN) graph. If provided, only correlations for pairs of cells connected in the KNN graph will be computed.

n_cores

The number of threads to use for parallel processing. Defaults to detectCores() - 1.

Value

A matrix of cell-to-cell correlation values.

Details

colDeltaCor_all calculates cell-to-cell correlations based on variance-stabilized transformed differences between observed and perturbed gene expression matrices. The function computes correlations either for all cell pairs or, when a KNN graph is provided, only for cells connected in the graph.

The primary steps of this analysis are:

1. Variance-stabilizing transformation: For each cell, observed and perturbed expression values are transformed to reduce variance effects.

2. Cell-to-cell correlation computation: For each cell, the function calculates correlation values with either all other cells or only the nearest neighbors, as defined by the KNN graph.

3. Parallel processing: Correlations are computed in parallel for optimized performance.