Test differential co-expression between two groups — coexpr

Uses the maintained Bioconductor diffcoexp implementation to identify differential links and genes from two condition-specific expression matrices. DCA here means differential co-expression analysis.

Usage

coexpr_differential(
  x,
  experiment,
  group,
  contrast,
  assay,
  features = NULL,
  correlation = "pearson",
  p_adjust = "BH",
  correlation_threshold = 0.5,
  correlation_fdr = 0.1,
  difference_threshold = 0.5,
  difference_fdr = 0.1,
  gene_fdr = 0.1,
  max_pairs = 5e+06
)

Arguments

x: A MultiAssayExperiment object.
experiment: Optional experiment name.
group: Metadata column or grouping vector.
contrast: Two group labels, in the order passed to exprs.1 and exprs.2.
assay: Assay name or one-based assay index.
features: Optional feature subset.
correlation: Correlation method.
p_adjust: Multiple-testing correction method.
correlation_threshold, correlation_fdr: Thresholds for condition-specific correlations.
difference_threshold, difference_fdr: Thresholds for changes in correlation.
gene_fdr: FDR threshold for differential co-expression genes.
max_pairs: Maximum number of feature pairs to analyze. Set to Inf only after considering the quadratic memory and runtime cost.

Value

The native result returned by diffcoexp::diffcoexp().

Details

Use a normalized, variance-stabilized or log-expression assay. Remove known unwanted variation before this analysis when appropriate. Correlations are estimated separately, so small groups are unstable even when they meet the minimum of four samples; groups below ten emit a warning.