profile_cluster_DEG — profile_cluster

Differential Expression Between Profile and Background in a Single Cluster

Usage

profile_cluster_DEG(
  profile,
  cluster,
  exprs,
  cell_meta,
  cell_profile_prob,
  cluster_col = "cluster",
  pseudotime_col = "cell_t",
  permute_n = 50
)

Arguments

profile: Character or numeric vector indicating which profiles to test.
cluster: Character or factor. Name or ID of the cluster to analyze.
exprs: A gene expression matrix (genes × cells).
cell_meta: A data frame of cell-level metadata (must include pseudotime and cluster labels).
cell_profile_prob: A numeric matrix of profile probabilities (cells × profiles).
cluster_col: Name of the column in `cell_meta` for cluster labels (default: `"cluster"`).
pseudotime_col: Name of the column in `cell_meta` representing pseudotime (default: `"cell_t"`).
permute_n: Integer. Number of permutations for null model generation (default: 50).

Value

A list containing:

stat: Data frame with statistics, empirical p-values, and FDR-adjusted p-values

cell: Character vector of cell IDs used in the test

design_null, design_full, coef, df: Model components from GAM fitting

Details

Performs pseudotime-aware differential expression testing between a target profile and control background within a specific transcriptional cluster. Combines soft cluster-weighted regression with permutation-derived null distributions to assess gene-level significance.

Examples

# Assuming cell_meta, exprs, and cell_profile_prob are defined
res <- profile_cluster_DEG("P1", cluster = "C2", exprs, cell_meta, cell_profile_prob)
#> Error in as.data.frame(cell_meta): object 'cell_meta' not found
head(res$stat)
#> Error in h(simpleError(msg, call)): error in evaluating the argument 'x' in selecting a method for function 'head': object 'res' not found