Create preprocessed input data for SCEG-HiC — process

This function generates preprocessed input data for the SCEG-HiC framework. The output can be directly used by calculateHiCWeights and Run_SCEG_HiC. It supports both aggregated (clustered) data and non-aggregated (cell-type-specific) data generation.

Usage

process_data(
  object,
  aggregate = TRUE,
  celltype = NULL,
  rna_assay = "RNA",
  atac_assay = "peaks",
  cellnames = NULL,
  atacbinary = TRUE,
  k_neigh = 50,
  max_overlap = 0.8,
  size_factor_normalize = TRUE,
  reduction.name = NULL,
  seed = 123,
  verbose = TRUE
)

Arguments

object: A Seurat object.
aggregate: Logical. Whether to generate aggregated (clustered) data. If FALSE, a specific celltype must be provided.
celltype: Character. A specific cell type to subset for non-aggregated data. Required if aggregate = FALSE.
rna_assay: Character. Name of the assay containing gene expression data. If aggregate = TRUE, this should contain raw expression counts. If aggregate = FALSE, this should contain normalized expression values.
atac_assay: Character. Name of the assay containing chromatin accessibility data. If aggregate = TRUE, this should contain raw accessibility counts. If aggregate = FALSE, this should contain normalized accessibility values.
cellnames: Character vector. Name(s) of one or more metadata columns used to group the cells. Default is the current cell identities.
atacbinary: Logical. Should the aggregated scATAC-seq matrix be binarized?
k_neigh: Integer. Number of cells to aggregate per group (default is 50).
max_overlap: Numeric. Maximum allowed overlap ratio between two aggregated groups (default is 0.8).
size_factor_normalize: Logical. Whether to normalize accessibility values using size factors.
reduction.name: Character. Name of the dimensionality reduction to use for extracting cell coordinates for aggregation.
seed: Integer. Random seed.
verbose: Logical. Should progress messages and warnings be printed?

Value

A list containing preprocessed data for SCEG-HiC:

If aggregate = TRUE, returns a list with aggregated RNA and ATAC matrices.
If aggregate = FALSE, returns RNA and ATAC matrices corresponding to the selected cell type.

Examples

data(multiomic_small)

# Aggregated data with paired scRNA-seq and scATAC-seq
SCEGdata <- process_data(multiomic_small, k_neigh = 5, max_overlap = 0.5)
#> Generating aggregated data
#> Aggregating cluster 0
#> Sample cells randomly.
#> There are 11 samples
#> Aggregating cluster 1
#> Sample cells randomly.
#> There are 11 samples

# Single-cell type data for cell type "1"
SCEGdata <- process_data(multiomic_small, aggregate = FALSE, celltype = "1")